Theses and Dissertations

2015

ABSTRACTS

Departamento de Informática 
Pontifícia Universidade Católica do Rio de Janeiro - PUC-Rio
Rio de Janeiro - Brazil
 

This file contains the list of the MSc. Dissertations and PhD. Thesis presented to the Departmento de Informática, Pontifícia Universidade Católica do Janeiro - PUC-Rio, Brazil, in 2015.  They are all available in print format and, according to the authors' preference, some of them are freely available for download, while others are freely available for download to the PUC-Rio community exclusively(*). 

For any requests, questions, or suggestions, please contact:
Rosane Castilho bib-di@inf.puc-rio.br

Last update: 21/DECEMBER/2016
 

INDEX

[In construction; sometimes digital versions may not be available yet]

[15_PhD_branco]
Adriano Francisco BRANCO. Scripting customized components for Wireless Sensor Networks. [Title in Portuguese: Programando redes de sensores sem fio com scripts sobre componentes customizados]. Ph.D. Thesis. Eng. Presentation: 10/09/2015. 100 p. Advisor: Noemi da La Roque Rodriguez.

Abstract: Programming wireless sensors networks (WSN) is a dicult task. The programmer must deal with several concurrent activities in an environment with severely limited resources. In this work we propose a programming model to facilitate this task. The model we propose combines the use of configurable component-based virtual machines with a reactive scripting language which can be statically analyzed to avoid unbounded execution and memory conflicts. This approach allows the exibility of remotely uploading code on motes to be combined with a set of guarantees for the programmer. The choice of the specific set of components in a virtual machine configuration defines the abstraction level seen by the application script. To evaluate this model, we built Terra, a system combining the scripting language Céu-T with the Terra virtual machine and a library of components. We designed this library taking into account the functionalities commonly needed in WSN applications - typically for sense and control. We implemented different applications using Terra and using an event-driven language based on C and we discuss the advantages and disadvantages of the alternative implementations. Finally, we also evaluate Terra by measuring its overhead in a basic application and discussing its use and cost in different WSN scenarios.

[15_PhD_skyrme]
Alexandre Rupert Arpini SKYRME. Safe record sharing in dynamic programming languages. [Title in Portuguese: Compartilhamento seguro de registros em linguages de programação dinâmicas]. Ph.D. Thesis. Eng. Presentation: 17/03/2015. 77 p. Advisors: Noemi de La Rocque Rodriguez and Roberto Ierusalimschy.

Abstract: Dynamic programming languages have become increasingly popular and have been used to implement a range of applications. Meanwhile, multicore processors have become the norm, even for desktop computers and mobile devices. Therefore, programmers must turn to parallelism as a means to improve performance. However, concurrent programming remains difficult. Besides, despite improvements in static languages, we find dynamic languages are still lacking in concurrency support. In this thesis, we argue that the main problem with concurrent programming is unpredictability unexpected program behaviors, such as returning out-of-thin-air values. We observe that unpredictability is most likely to happen when shared memory is used. Consequently, we propose a concurrency communication model to discipline shared memory in dynamic languages. The model is based on the emerging concurrency patterns of not sharing data by default, data immutability, and types and effects (which we turn into capabilities). It mandates the use of shareable objects to share data. Besides, it establishes that the only means to share a shareable object is to use message passing. Shareable objects can be shared as read-write or read-only, which allows both individual read-write access and parallel read-only access to data. We implemented a prototype in Lua, called luashare, to experiment with the model in practice, as well as to carry out a general performance evaluation. The evaluation showed us that safe data sharing makes it easier to allow for communication among threads. Besides, there are situations where copying data around is simply not an option. However, enforcing control over shareable objects has a performance cost, in particular when working with nested objects.

[15_MSc_villarmosa]
Alexandre Menezes VILLARMOSA. Modelagem de agentes computacionais baseada na doutrina militar. [Title in English: Computational modeling agents based on military doctrine]. M.Sc. Diss. Port. Presentation: 14/07/2015. 84 p. Advisor: Marcelo Gattass.

Abstract: Since the beginning of nineteenth century combat simulations are used in military training. It’s necessary to involve lots of military to these trainings occur reliably. In the late 1940s the idea of computational agents was developed in artificial intelligence and showed as an excellent tool to reduce the amount of personnel involved in combat simulations. Agents perceive the environment where they are inserted and take actions upon it following a set of rules. That reminds the behavior of a soldier. A soldier, or a group of then, perceive the battlefield and take a series of actions based on military doctrine. Therefore, the scope of this work is to present a viable way to define the behavior of computational agents based on military doctrine, so that they can replace some of the personnel involved in a combat simulation without affecting the reliability of the training in course. In addition making more efficient simulation systems, reducing the amount of required military for its proper implementation, can also help to check the logical consistency of the actions planned in the doctrinal manuals.

[15_MSc_moreira]
André de Souza MOREIRA. Engenharia reversa em modelos CAD utilizando descritores de forma e máquina de vetores de suporte. [Title in English: Reverse engineering for CAD models using shape descriptors and support vector machine]. M.Sc. Diss. Port. Presentation: 27/03/2015. 72 p. Advisor: Marcelo Gattass.

Abstract: 3D CAD Models have played an important role in engineering projects’ management. It is noticeable in many of these files the presence of several objects with implicit representation that end up being represented as triangular meshes. Although suitable for rendering, the triangular mesh representation brings some drawbacks, such as the ambiguity in objects with low discretization rate. The reverse engineering aims to reconstruct this discrete representation to its original continuous representation. In this work, we propose a novel methodology for geometry reconstruction in CAD models using Support Vector Machines and Shape Descriptors.

[15_PhD_maidl]
André Murbach MAIDL. Typed Lua: an optional type system for LUA. [Title in English: Typed Lua: um sistema de tipos opcional para Lua]. Ph. D. Thesis. Port. Presentation: 10/04/2015. 149 p. Advisors: Roberto Ierusalimschy and Fábio Mascarenhas de Queiroz (UFRJ).

Abstract: Dynamically typed languages such as Lua avoid static types in favor of simplicity and exibility, because the absence of static types means that programmers do not need to bother with abstracting types that should be validated by a type checker. In contrast, statically typed languages provide the early detection of many bugs, and a better framework for structuring large programs. These are two advantages of static typing that may lead programmers to migrate from a dynamically typed to a statically typed language, when their simple scripts evolve into complex programs. Optional type systems allow combining dynamic and static typing in the same language, without affecting its original semantics, making easier this code evolution from dynamic to static typing. Designing an optional type system for a dynamically typed language is challenging, as it should feel natural to programmers that are already familiar with this language. In this work we present and formalize the design of Typed Lua, an optional type system for Lua that introduces novel features to statically type check some Lua idioms and features. Even though Lua shares several characteristics with other dynamically typed languages such as JavaScript, Lua also has several unusual features that are not present in the type system of these languages. These features include functions with exible arity, multiple assignment, functions that are overloaded on the number of return values, and the incremental evolution of record and object types. We discuss how Typed Lua handles these features and our design decisions. Finally, we present the evaluation results that we achieved while using Typed Lua to type existing Lua code.

[15_PhD_monteiro]
Andrei Alhadeff MONTEIRO. Mapping cohesive fracture and fragmentation simulations to GPUs. [Title in Portuguese: Compartilhamento seguro de registros em linguages de programação dinâmicas]. Ph.D. Thesis. Eng. Presentation: 16/09/2015. 156 p. Advisor: Waldemar Celes Filho.

Abstract: A GPU-based computational framework is presented to deal with dynamic failure events simulated by means of cohesive zone elements. We employ a novel and simplified topological data structure relative to CPU implementation and specialized for meshes with triangles or tetrahedra, designed to run efficiently and minimize memory requirements on the GPU. We present a parallel, adaptive and distributed explicit dynamics code that implements an extrinsic cohesive zone formulation where the elements are inserted “on-the-fly”, when needed and where needed. The main challenge for implementing a GPU-based computational framework using an extrinsic cohesive zone formulation resides on being able to dynamically adapt the mesh, in a consistent way, by inserting cohesive elements on fractured facets and inserting or removing bulk elements and nodes in the adaptive mesh modification case. We present a strategy to refine and coarsen the mesh to handle dynamic mesh modification simulations on the GPU. We use a reduced scale version of the experimental specimen in the adaptive fracture simulations to demonstrate the impact of variation in floating point operations on the final fracture pattern. A novel strategy to duplicate ghost nodes when distributing the simulation in different compute nodes containing one GPU each is also presented. Results from parallel simulations show an increase in performance when adopting strategies such as distributing different jobs amongst threads for the same element and launching many threads per element. To avoid concurrency on accessing shared entities, we employ graph coloring for non-adaptive meshes and node traversal for the adaptive case. Experiments show that GPU efficiency increases with the number of nodes and bulk elements

[15_MSc_lutfi]
Antonio Nascimento LUTFI. Um framework para game shows interativos de TV com realidade aumentada e segunda tela. [Title in English: A framework for TV game shows with second screen]. M.Sc. Diss. Port. Presentation: 17/04/2015. 94 p. Advisor: Bruno Féijo.

Abstract: Presents a framework for the development of interactive TV game shows using augmented reality in TV studios, which allows the participation of viewers using tablets and smartphones as the second screen. This research also investigates new convergence paradigms between TV and video games.

[15_MSc_escobarendara]
Ariel ESCOBAR ENDARA. CubMed: um framework para a criação de aplicações de assistência médica ubíqua baseado em agentes de software colaborativos. [Title in English: CubMed: a framework for the creation of ubiquitous medical assistance applications based on collaborative software agents]. M.Sc. Diss. Port. Presentation: 28/08/2015. 90 p. Advisor: Carlos José Pereira de Lucena.

Abstract: The health area needs to deal with various problems related to issues of infrastructure, lack of qualified personnel and a large number of patients. As a solution to problems of this nature, u-Healthcare was created as an application of the concepts of Ubiquitous Computing (UbiComp) in the area of health care. u-Healthcare allows health monitoring at any time and place from electronic devices connected to the Internet. However, the expansion of health monitoring for an ubiquitous environment cannot be performed with protocols and procedures currently used, since this approach would drastically increase the consumption of time and resources. For that reason, the development of tools to provide health services can be supported in research areas such as Multi-Agent System (MAS) and Computer Supported Cooperative Work (CSCW). In that sense, MAS can be used to automate processes through the properties of software agents. On the other hand CSCW gives the possibility of establishing a model of cooperation among the participants on the application. Based on these aspects, this work proposes the modeling and development of a framework capable of providing support and help on the construction of dedicated u-Healthcare applications which should be based on the concepts of MAS and CSCW. To illustrate the use of the framework, there are presented two scenarios of use. The first scenario corresponds to a fetal monitoring system, which allows early detection of fetal abnormalities. The second scenario consists of a drug administration assistant, which allows the doctor to control drug use by his patients.

[15_PhD_baffa]
Augusto Cesar Espídola BAFFA. Storytelling based on audience social interaction. [Title in Portuguese: Storytelling baseado na interação social da audiência]. Ph.D. Thesis. Port. Presentation: 07/07/2015. 91 p. Advisor: Bruno Féijo.

Abstract: To tell a story, the storyteller uses all his/her skills to entertain an audience. This task not only relies on the act of telling a story, but also on the ability to understand reactions of the audience during the telling of the story. It is not so difficult to adapt a story for a single individual based on his/her preferences and previous choices. However, the task of choosing what is best for a group becomes quite complicated. The selection by majority voting cannot be effective because it can discard alternatives that are secondary for some individuals, but that would work better for the group in question. Thus, the careless selection of events in a story could cause audience splitting, causing some people to give up keep watching because they were not pleased. This thesis proposes a new methodology to create tailored stories for an audience based on personality traits and preferences of each individual. As an audience may be composed of individuals with similar or mixed preferences, it is necessary to consider a middle ground solution based on the individual options. In addition, individuals may have some kind of relationship with others who influence their decisions. The proposed model addresses all steps in the quest to please the audience. It infers what the preferences are, computes the scenes reward for all individuals, estimates their choices independently and in group, and allows Interactive Storytelling systems to find the story that maximizes the expected audience reward. The proposed model can easily be extended to other areas that involve users interacting with digital environments.

[15_PhD_figueiredo]
Aurélio Moraes FIGUEIREDO. Mapeamento de eventos sísmicos baseado em algoritmos de agrupamento de dados. [Title in English: Mapping seismic events using clustering-based methodologies]. Ph.D. Thesis. Port. Presentation: 20/08/2015. 69 p. Advisor: Marcelo Gattass.

Abstract: We present clustering-based methodologies used to process 3D seismic data. It firstly replaces the volume voxels by corresponding feature samples representing the local behavior in the seismic trace. After this step samples are used as entries to clustering procedures, and the resulting cluster maps are used to create a new representation of the original volume data. This strategy finds the global structure of the seismic signal. It strongly reduces the impact of noise and small disagreements found in the voxels of the entry volume. These clustered versions of the input seismic data can then be used in two different applications: to map 3D horizons automatically and to produce visual attribute volumes where seismic faults and any discontinuities present in the data are highlighted. Concerning the horizon mapping, as the method does not use any lateral similarity measure to organize horizon voxels into clusters, the methodology is very robust when mapping difficult cases. It is capable of mapping a great portion of the seismic interfaces present in the data. In the case of the visualization attribute, it is constructed by applying an auto-adaptable function that uses the voxel neighboring information through a specific measurement that globally highlights the fault regions and other discontinuities present in the original volume. We apply the methodologies to real seismic data, mapping even seismic horizons severely interrupted by various discontinuities and presenting visualization attributes where discontinuities are adequately highlighted.

[15_MSc_pedras]
Bernardo Frankenfeld Villela PEDRAS. EnvironRC: integrating collaboration and mobile communication to offshore engineering virtual reality applications. [Title in Portuguese: EnvironRC: integrando colaboração e comunicação móvel a aplicações de Engenharia Off-Shore em ambiente de realidade virtual]. M.Sc. Diss. Eng. Presentation: 11/09/2015. 90 p. Advisor: Alberto Barbosa Raposo.

Abstract: Offshore Engineering visualization applications are, on most cases, very complex and should display a lot of data coming from very computational intensive numerical simulations. To help analyze and better visualize the results, 3D visualization can be used in conjunction with a Virtual Reality (VR) environment. The main idea for this work began as we realized two different demands that engineering applications had when running on VR setups: firstly, a demand for visualization support in the form of better navigation and better data analysis capabilities. Secondly, a demand for collaboration, due to the difficulties of coordinating a team with one member using VR. To meet this demands, we developed a Service Oriented Architecture (SOA) capable of adding collaboration capabilities to any application. The idea behind our solution is to enable real-time data visualization and manipulation on tablets and smartphones. Such devices can be used to help navigate the virtual world or be used as a second screen, helping visualize and manipulate large sets of data in the form of tables or graphs. Furthermore, we want to allow collaboration-unaware application to collaborate with as little reworking of the original application as possible. Another big advantage that mobile devices bring to the engineering applications is the capability of accessing the data on remote locations, like on oil platforms or refineries, and so allowing the field engineer to check the data or even change it on the fly. As our test application, we used ENVIRON, which is a VR application for visualization of 3D models and simulations developed in collaboration with a team from the Institute Tecgraf of PUC-Rio. We added this solution to ENVIRON and it was tested with an experiment and during a review process of Offshore Engineering using VR Setups (Power wall and CAVE).

[15_MSc_cruz]
Breno Riba da Costa CRUZ. Uma interface de programação para controle de sobrecarga em arquiteturas baseadas em estágios. [Title in English: A programming interface for overload control in staged event based architectures]. M.Sc. Diss. Port. Presentation: 24/02/2015. 75 p. Advisors: Noemi da La Roque Rodriguez and Ana Lúcia de Moura.

Abstract: Specific scheduling policies can be appropriate for overload control in different application scenarios. However, these policies are often difficult to implement, leading developers to reprogram entire systems in order to adapt them to a particular policy. Through the study of various  scheduling policies, we propose an interface model that allows the programmer to integrate new policies and monitoring schemes to the same application in a Staged Event-Driven Architecture. We describe the implementation of the proposed interface and the results of it's use in implementing a set of scheduling policies for two applications with different load profiles.

[15_MSc_chagas]
Bruno Azevedo CHAGAS. End-user configuration in assistive technologies: a case study with a severely physically impaired user. [Title in Portuguese: Configuração pelo usuário final em tecnologias assistivas: um estudo de caso com um usuário com limitação física severa]. M.Sc. Diss. Port. Presentation: 04/09/2015. 124 p. Advisors: Hugo Fuks and Clarisse Sieckenius de Souza.

Abstract: Assistive Technology (AT) aims at compensating for motor, sensory or cognitive functional limitations of its users. One of the reasons AT is hard to design and turn into a product is the variability of kinds and degrees of disabilities and individual characteristics among users (physical, psychological, cultural and environmental). This variability can be addressed by means of configurations. This work takes as a starting point the premise that the ability for the end-user to adapt AT may have the potential to improve user’s experience and the quality of the products. However, before engaging in such endeavor we must answer questions like: what is configuration in the AT domain? What does AT mean to users (and stakeholders)? What could, should or should not be configured and how? In this work, we conducted a case study mixing ethnography and action research with a single tetraplegic participant who came to our lab seeking for technology that could help him in his daily life. First, we interviewed him and observed his daily needs and activities and then we developed an AT platform prototype that controls some devices to be operated simultaneously by gesture and voice interaction in “his smart home.” Throughout two action-research cycles, we investigated interaction and technological issues regarding our prototype configuration and use. Based on our findings, we propose a set of dimensions and a collaborative framework for AT configuration. Our main contribution is to propose a conceptual structure for organizing the AT configuration problem space to support the design of similar technologies.

[15_PhD_cafeo]
Bruno Barbieri de Pontes CAFEO. On the relationship between feature dependencies and change propagation. [Title in Portuguese: Investigando o relacionamento entre dependências de características e progação de mudanças]. Ph. D. Thesis. Eng. Presentation: 12/06/2015. 167 p. Advisor: Alesandro Fabricio Garcia.

Abstract: Features are the key abstraction to develop and maintain software product lines. A challenge faced in the maintenance of product lines is the understanding of the dependencies that exist between features. In the source code, a feature dependency occurs whenever program elements within the boundaries of a feature’s implementation depend on elements external to that feature. Examples are either attributes or methods defined in the realisation of a feature, but used in the code realising other features. As developers modify the source code associated with a feature, they must ensure that other features are consistently updated with the new changes – the so-called change propagation. However, appropriate change propagation is far from being trivial as features are often not modularised in the source code. In this way, given a change in a certain feature, it is challenging to reveal which (part of) other features should also change. Change propagation becomes, therefore, a central and non-trivial aspect of software product-line maintenance. Developers may overlook important parts of the code that should be revised or changed, thus not fully propagating changes. Conversely, they may also unnecessarily analyse parts that are not relevant to the feature-maintenance task at hand, thereby increasing the maintenance effort or even mis-propagating changes. The creation of a good mental model based on the structure of feature dependencies becomes essential for gaining insight into the intricate relationship between features in order to properly propagate changes. Unfortunately, there is no understanding in the state of the art about structural properties of feature dependencies that affect change propagation. This understanding is not yet possible as: (i) there is no conceptual characterisation and quantification means for structural properties of feature dependency, and (ii) there is no empirical investigation on the influence of these properties on change propagation. In this context, this thesis presents three contributions to overcome the aforementioned problems. First, we develop a study to understand change propagation in presence of feature dependencies in several industry-strength product lines. Second, we propose a measurement framework intended to quantify structural properties of feature dependencies. We also develop a study revealing that conventional metrics typically used in previous research, such as coupling metrics, are not effective indicators of change propagation in software product lines. Our proposed metrics consistently outperformed conventional metrics. Third, we also propose a method to support change propagation by facing the organisation of feature dependency information as a clustering problem. We evaluate if our proposed organisation has potential to help developers to propagate changes in software product lines.

[15_MSc_amaral]
Bruno Guberfain do AMARAL. A visual analysis of bus GPS data in Rio. [Title in Portuguese: Uma análise visual dos dados de GPS dos ônibus no Rio]. M.Sc. Diss. Port. Presentation: 12/06/2015. 45 p. Advisor: Hélio Cortes Vieira.

Abstract: Smart cities is a current subject of interest for public administrators and researchers. Getting the cities smarter is one of the challenges for the near future, due to the growing demand for public services. In particular, public transportation is one of most visible aspects of a living city and, therefore, its implementation must be very efficient. The public transportation system of the City of Rio de Janeiro is historically deficient, mostly because it is based on an old bus system. To change it, the City Hall took some actions, such as the development of an open data project that shows, at about every minute, the GPS instant position of all buses in the city. Although it is not a new technology, it is the first initiative to be developed in Rio. This work presents simple tools for the visual exploration of this big dataset based on the historical information from this service, which reaches a total of more than one billion samples. With these tools one is able to discover trends, identify patterns, and locate abnormalities within the massive collection of the buses' GPS data.

[15_MSc_souza]
Bruno José Olivieri SOUZA. An approach for movement coordination of swarms of unmanned aerial vehicles using mobile networks. [Title in Portuguese: Uma abordagem para a coordenação do movimento de enxames de veículos aéreos não tripulados usando redes móveis]. M.Sc. Diss. Port. Presentation: 18/03/2015. 67 p. Advisor: Markus Endler.

Abstract: This work presents an approach to coordinate swarms of Unmanned Aerial Vehicles (UAV) based on Internet communication provided by mobile phone networks. Several activities can be done by several UAVs flying in formation, such as surveillance and monitoring of mass events, search and rescue tasks, control of agricultural pests, monitoring and forest conservation, inspection of pipelines and electricity distribution networks or even military attack and recognition missions. Coordination of UAVs swarm can be branch in two sub-problems: communication between members of the swarm and the algorithm that controls members’ behaviors regarding their movements. The proposed solution assumes the use of a smartphone coupled with each UAV of the swarm, in order to provide the required level of reliable communication on the mobile Internet and run the proposed algorithm for the coordination of swarms of UAVs. Experiments were performed with emulated UAVs and WAN mobile networks. The results have demonstrated the effectiveness of the proposed algorithm, and have shown the influence of the network latency and the UAV speeds on the accuracy of the movement coordination in the swarms.

[15_MSc_vanderley]
Carla Galdino VANDERLEY. Uma sistemática de monitoramento de erros em sistemas distribuídos. [Title in English: A systematic for error monitoring in distributed systems]. M.Sc. Diss. Port. Presentation: 01/04/2015. 56 p. Advisor: Arndt von Staa.

Abstract: Systems formed by distributed components enable the occurrence of faults arising from the interaction between components. The stage of testing this class of systems is difficult, since foresee all interactions between components of a system is not feasible. Therefore, even if a component system is tested, the occurrence of run-time errors is still possible and, of course, these errors should be seen shooting some action that prevents them from causing major damage. This work presents an identification mechanism of errors given by semantic inconsistency typed data, based on structured logs. Semantic inconsistency of typed data could cause failures due to the misinterpretation of values that are represented syntactically under the same basic types. The proposed mechanism consists in generating structured logs according to the definition of communication interfaces, and identifying anomalies by an existing contract verification technique. In addition, the mechanism proposes a management model of taxonomic semantic types using the Case-Based reasoning technique (CBR). The fault identification mechanism was implemented through an extension of the Robot Operation System middleware (ROS). The mechanism, in addition to observing errors, generates additional information which aims to assist the diagnosis of the cause of the observed error. Finally, a proof of concept applied to a locomotion control system for a hybrid robot, adapted from a real system, has been developed for fault identification validation.

[15_MSc_fernandes]
Chrystinne Oliveira FERNANDES. IoT4Health: um framework no domínio de e-health para acompanhamento de pacientes ulilizando agentes de software. [Title in English: IoT4Health: a framework in e-health domain for patiente monitoring by using sofware agents]. M.Sc. Diss. Port. Presentation: 27/08/2015. 144 p. Advisor: Carlos José Pereira de Lucena.

Abstract:  The search for innovative solutions in E-Health domain has largely motivated the conduct of scientific research in this area, whose exploration can bring numerous benefits to society. Despite the technological resources available nowadays, there are still many problems in the hospital environment. Aiming contribute to the development of technological solutions applied to this area, we highlighted the development of the IoT4Health framework. Two instances of this framework were built to serve as concept proof 1-Agents4Health; 2-Remote Patient Monitoring (RPM); 3-EHealth System. The Agents4Health is a multi-agent system in E-health domain supported by an Internet of Things (IoT) solution to automate techniques commonly used in patients' treatment and data collection processes. This solution comprises software agents and hardware prototypes including sensors, micro-controllers that work together to make hospital environments more proactive. In addition, the solution provides remote storage of patient data in cloud-based platforms, allowing external professionals to work collaboratively with the local team. A Web system enables real-time visualization of patient's record captured through sensors, such as temperature and heart rate values displayed as graphical charts through an intuitive interface. Software agents constantly monitor collected data to detect anomalies in patients' health status and send alerts to health professionals. The RPM also supports patient monitoring acivities by using mobile applications, with focus on patient evolution. Finally, the EHealth system comprises the set of applicaions created in order to validate our tool.

[15_MSc_gama]
Daltro Simões GAMA. Integração de um sistema de submissão batch com um mmbiente de computação em nuvem. [Title in English: Integration of a batch subimission system with a cloud computing environment]. M.Sc. Diss. Port. Presentation: 21/12/2015. 83 p. Advisors: Noemi de La Rocque Rodriguez and Maria Júlia de Lima.

Abstract: Cloud computing appeals to those who need many machines to run their programs, attracted by low maintenance costs and easy configuration. In this work we implemented a new integration for the CSGrid system, from Tecgraf/PUC-Rio, enabling it to submit workloads to Microsoft Azure public cloud, thus enjoying the benefits of elastic computing resources. For this purpose, we present related works and some performance measures in the case of CSGrid's use of Microsoft Azure public cloud, with regard to costs on data transfers and provisioning of virtual machines. With this integration, we could evaluate the benefits and difficulties involved in using cloud resources in a system designed for the submission of HPC applications to clusters.

[15_MSc_marques]
Daniel da Rosa MARQUES. Um metaclassificador para encontrar k-classes mais relevantes. [Title in English: A metaclassifier for finding the k-classes most relevants]. M.Sc. Diss. Port. Presentation: 24/11/2015. 56 p. Advisor: Eduardo Sany Laber.

Abstract: Consider a network with k nodes that may fail along its operation. Furthermore assume that it is impossible to check all nodes whenever a failure occurs. Motivated by this scenario, we propose a method that uses supervised learning to generate rankings of the most likely nodes responsible for the failure. The proposed method is a meta-classifier that is able to use any kind of classifier internally, where the model generated by the meta-classifier is a composition of those generated by the internal classifiers. Each internal model is trained with a subset of the data created from the elimination of instances whose classes were already put in the ranking. Metrics derived from Accuracy, Precision and Recall were proposed and used to evaluate this method. Using a public data set, we verified that the training and classification times of the meta-classifier were greater than those of a simple classifier. However it reaches better results in some cases, as with the decision trees, that exceeds the benchmark accuracy for a margin greater than 5%.

[15_MSc_ferreira]
Daniel Vitor Costa FERREIRA. Lean communication-centered design: um processo leve de design centrado na comunicação. [Title in English: Lean communication-centred design: a lightweight design process]. M.Sc. Diss. Port. Presentation: 10/09/2015. 159 p. Advisor: Simone Diniz Junqueira Barbosa.

Abstract: Lean Communication-Centered Design (LeanCCD) is a Human-Computer Interaction (HCI) design process, which consists of conducting a workshop, detailing user goals, combining interaction models with paper sketches, and testing them with users, supported by guides and templates. This study adapted the Communication-Centered Design (CCD) and the eXtreme Communication-Centered Design (eXCeeD), other communication-centered design processes grounded on Semiotic Engineering (SemEng). SemEng defines the interaction as a computer-mediated communication process between designers and users. Approaches and processes based on SemEng are not used to directly yield the answer to a problem, but to increase the problem-solver’s understanding of the problem itself and the implication it brings about. Process evaluation in a case study, in the industry, proved itself difficult, both in carrying out LeanCCD activities and in the correct application of some techniques and concepts. However, unlike eXCeeD, we were able to observe a systematic use of questions that contributed to designers’ reflection, aided by the proposed templates and guides.

[15_MSc_silva]
Davidson Felipe da SILVA. Keepfast – um ambiente extensível dirigido por modelos para avaliação de desempenho de páginas web. [Title in English: Keepfast - a model-driven framework for the evaluation web page performance]. MSc. Diss. Port. Presentation: 03/09/2015. 76 p. Advisor: Daniel Schwabe.

Abstract: Over the years, there has been an increase in complexity on the client side of Web application architecture. Consequently, the addition of functionality or change in the implementations on the client side often leads to a drop in performance, which should be avoided. This effect is compounded due to the constant evolution of implementation technologies and the growing number of devices with web connection. This work presents a Model-driven framework for the assessing web page performance, in the application client side, allowing for a variety of performance evaluation contexts. This environment can be extended and customized to reflect the most important features that the designer wants to evaluate. A case study is presented, showing the results obtained in a real scenario compared to other methodologies available.

[15_PhD_barbosa]
Eiji Adachi Medeiros BARBOSA. Global-aware recommendations for repairing exception handling violations. [Title in Portuguese: Recomendações globais para reparaçâo de violações de tratamento de exceções]. MSc. Diss. Eng. Presentation: 00/11/2015. 213 p. Advisor: Alessandro Fabrício Garcia.

Abstract: Exception handling is the most common way of dealing with exceptions in robust software development. Exception handling refers to the process of signaling exceptions upon the detection of runtime errors and taking actions to respond to their occurrence. Despite being aimed at improving software robustness, software systems are still implemented without relying on explicit exception handling policies. Each policy defines the design decisions governing how exception handling should be implemented in a system. These policies are often not documented and are only implicitly defined in the system design. Thus, developers tend to introduce in the source code violations of implicit policies and these violations commonly cause failures in software systems. In this context, the goal of this thesis is to support developers in detecting and repairing exception handling violations. To achieve this goal, two complementary solutions were proposed. The first solution is based on a domain-specific language supporting the detection of violations by explicitly defining exception handling policies to be enforced in the source code. The proposed language was evaluated with a user-centric study and a case study. With the observations and experiences gathered in the user-centric study, we identified some language characteristics that hindered its use and that motivated new language constructs. In addition, the results of the case study showed that violations and faults in exception handling share common causes. Therefore, violations can be used to detect potential causes of exceptionrelated failures. To complement the detection of exception handling violations, this work also proposed a solution for supporting the repair of exception handling violations. Repairing these violations requires reasoning about the global impact that exception handling changes might have in different parts of the system. Thus, this work proposed a recommender heuristic strategy that takes into account the global context of where violations occur to produce recommendations. Each recommendation produced consists of a sequence of modifications that serves as a detailed blueprint of how an exception handling violation can be removed from the source code. The proposed recommender strategy also takes advantage of explicit policy specifications, although their availability is not mandatory. The results of our empirical assessments revealed that the proposed recommender strategy produced recommendations able to repair violations in approximately 70% of the cases. When policy specifications are available, it produced recommendations able to repair violations in 97% of the cases.

[15_MSc_menendez]
Elisa Souza MENENDEZ. Materialized sameAs link maintenance with views. [Title in English: Manutenção de links sameAs materializados utilizando visões]. MSc. Diss. Port. Presentation: 20/07/2015. 68 p. Advisor: Marco Antonio Casanova.

Abstract: In the Linked Data field, data publishers frequently materialize sameAs links between two different datasets using link discovery tools. However, it may be difficult to specify linking conditions, if the datasets have complex models. A possible solution lies in stimulating dataset administrators to publish simple predefined views to work as resource catalogues. A second problem is related to maintaining materialized sameAs linksets, when the source datasets are updated. To help solve this second problem, this work presents a framework for maintaining views and linksets using an incremental strategy. The key idea is to re-compute only the set of updated resources that are part of the view. This work also describes an experiment to compare the performance of the incremental strategy with the full re-computation of views and linksets.

[15_PhD_leal]
Ericsson de Souza LEAL. Traçado de linhas de fluxo em modelos de reservatórios naturais de petróleo baseado em métodos numéricos adaptativos. [Title in English: Streamline tracing for oil natural reservoirs based on adaptive numerical methods]. MSc. Diss. Port. Presentation: 01/09/2015. 51 p. Advisor: Waldemar Celles Filho.

Abstract: Traditionally, streamlines in discrete models of natural oil reservoirs are traced by solving an ordinary differential equation in a cell-by-cell way, using analytical or numerical solutions, considering the local velocity of each cell. This strategy has a disadvantage: the streamline is traced considering a discrete, and so discontinuous, vector field. Furthermore, for massive models, to solve the equation in a cell-by-cell way may be inefficient. In this work, we explore a different strategy: the streamline tracing considers a continuous vector field represented by the discrete model. Therefore, we propose: (i) to use a spatial structure to speed up the point location process inside the reservoir model; (ii) to use spherical interpolation to obtain the velocity field from the discrete model; (iii) to use an adaptive numerical method to control the numerical error from the integration process. The results obtained for actual reservoir models demonstrate that the proposed method fulfills the precision requirements, keeping a good performance.

[15_PhD_sacramentoferreira]
Eveline Russo SACRAMENTO FERREIRA. An approach for dealing with inconsistencies in data mashups. [Title in Portuguese: Uma abordagem para lidar com inconsistências em combinações da dados]. MSc. Diss. Eng. Presentation: 11/09/2015. 100 p. Advisor: Marco Antonio Casanova.

Abstract: With the amount of data available on the Web, consumers can “mashup”and quickly integrate data from different sources belonging to the same application domain. However, data mashups constructed from independent and heterogeneous data sources may contain inconsistencies and, therefore, puzzle the user when observing the data. This thesis addresses the problem of creating a consistent data mashup from mutually inconsistent data sources. Specifically, it deals with the problem of testing, when data to be combined is inconsistent with respect to a predefined set of constraints. The main contributions of this thesis are: (1) the formalization of the notion of consistent data mashups by treating the data returned from the data sources as a default theory and considering a consistent data mashup as an extension of this theory; (2) a model checker for a family of Description Logics, which analyzes and separates consistent from inconsistent data and also tests the consistency and completeness of the obtained data mashups; (3) a heuristic procedure for computing such consistent data mashups.

[15_PhD_guimaraes]
Everton Tavares GUIMARÃES. A blueprint-based approach for prioritizing and ranking critical code anomalies. [Title in Portuguese: Uma abordagem baseada em blueprints para priorização e classificação de anomalias de código críticas. MSc. Diss. Port. Presentation: 12/09/2015. 142 p. Advisor: Alessandro Fabrício Garcia.

Abstract: Software systems are often evolving due to many changing requirements. As the software evolves, it grows in size and complexity, and consequently, its architecture design tends to degrade. Architecture degradation symptoms are often a direct consequence of the progressive insertion of code anomalies in the software implementation. A code anomaly is a recurring implementation structure that possibly indicates deeper architectural design problems. Code anomaly is considered critical when it is related with a structural problem in the software architecture. Its criticality stems from its negative influence on a wide range of non-functional requirements. For instance, the presence of critical code anomalies hinders software maintainability, i.e. these critical anomalies require wide refactoring in order to remove an architectural problem. Symptoms of architecture degradation have often to be observed in the source code due to the lack of an explicit, formal representation of the software architecture in a project. Many approaches are proposed for detecting code anomalies in software systems, but none of them efficiently support the prioritization and ranking of critical code anomalies according to their architecture impact. Our work investigates how the prioritization and ranking of such critical code anomalies could be improved by using blueprints. Architecture blueprints are usually provided by software architects since the early stages of the system development. Blueprints are informal design models usually defined to capture and communicate key architectural design decisions. Even though blueprints are often incomplete and inconsistent with respect to the underlying implementation, we aim to study if their use can contribute to improve the processes of prioritizing and ranking critical code anomalies. Aiming to address these research goals, a set of empirical studies has been performed. We also proposed and evaluated a set of heuristics to support developers when prioritizing and ranking code anomalies in 3 software systems. The results showed an average accuracy higher than 60% when prioritizing and ranking code anomalies associated with architectural problems in these systems.

[15_MSc_bertti]
Ezequiel BERTTI. MIRA – um ambiente para interfaces dirigidas por modelos para aplicações REST. [Title in Portuguese: MIRA – a model-driven interface framework for REST applications]. MSc. Diss. Port. Presentation: 02/03/2015. 137 p. Advisor: Daniel Schwabe.

Abstract: This work presents a Model-driven framework for the design of interfaces for REST applications. The framework allows building interfaces with minimal programming, The models used, as well as the generated interfaces are represented using W3C standards. A qualitative evaluation indicates that there are gains in both productivity and quality of the generated interfaces, when compared with traditional approaches.

[15_PhD_silva]
Fabio Araujo Guilherme da SILVA. Emotions in plots with non-deterministic planning for interactive storytelling. [Title in Portuguese: Emoções em enredos com planejamento não-determinístico para narração interativa]. Ph.D. Thesis. Port. Presentation: 14/04/2015. 156 p. Advisor: Antonio L. Furtado.

Abstract: Interactive storytelling is a form of digital entertainment in which users participate in the process of composing and dramatizing a story. In this context, determining the characters’ behaviour according to their individual preferences can be an interesting way of generating plausible stories where the characters act in a believable manner. Diversity of stories and opportunities for interaction are key requirements to be considered when designing such applications. This thesis proposes the creation of an architecture and a prototype for the generation and dramatization of interactive nondeterministic plots, using a model of emotions that not only serves to guide the actions of the characters presented by the plan generation algorithm, but also influences the participation of the users. Also, to improve the quality and diversity level of the stories, characters must be able to evolve in terms of their personality traits as the plot unfolds, as a reflection of the events they perform or are exposed to.

[15_MSc_moreira]
Felipe Baldino MOREIRA. An experiment on conceptual design of pervasive mobile games using quality requirements. [Title in Portuguese: Um experimento no design conceitual de jogos pervasivos móveis usando requisitos de qualidade]. MSc. Diss. Port. Presentation: 10/07/2015. 88 p. Advisor: Bruno Feijó.

Abstract: Pervasive games is an emerging game genre that mixes up mobile devices (such as smartphones), virtual worlds, and gameplay based on the real world, creating a mixed-reality game. This recent area lacks literature about conceptual design and quality requirements related to pervasiveness – the ultimate and elusive quality that differentiate pervasive mobile games from traditional digital games. In the present work, we discuss the development of a pervasive mobile game using quality requirements related to pervasiveness. Also, we consider those requirements in the entire game project (e.g., design, production, and post-production stages), focusing on the analysis, implementation, and gameplay. We expect that our results could help in improving the current state of design guidelines to develop pervasive mobile games.

[15_MSc_ferreira]
Fischer Jônatas FERREIRA. Uma análise da eficácia de assertivas executáveis como observadora de falhas em software. [Title in Portuguese: An effective analysis of executable assertives as indicators of software fails]. MSc. Diss. Port. Presentation: 09/04/2015. 117 p. Advisor: Arndt von Staa.

Abstract: Absolute reliability of software is considered unattainable, because even when it is build following strict quality rules, software is not free of failure occurrences during its lifetime. Software’s reliability level is related, among others, to the amount of remaining defects that will be exercised during its use. If software contains less remaining defects, it is expected that failures will occur less often, although many of these defects will never be exercised during its useful life. However, libraries and remote services of dubious quality are frequently used. In an attempt to enable software to check mistakes at runtime, hypothetically Lightweight Formal Methods, by means of executable assertions, can be effective and economically viable to ensure software’s reliability both at test time as well as at run-time. The main objective of this research is to evaluate the effectiveness of executable assertions for the prevention and observation of run-time failures. Effectiveness was evaluated by means of experiments. We instrumented data structures with executable assertions, and subjected them to tests based on mutations. The results have shown that all non-equivalent mutants were detected by assertions, although several of them were not detected by tests using non-instrumented versions of the programs. Furthermore, estimates of the computational cost for the use of executable assertions are presented. Based on the infrastructure created for the experiments we propose an instrumentation policy using executable assertions to be used for testing and to safeguard run-time.

[15_MSc_cunha]
Francisco José Plácido CUNHA. JAT4BDI: uma nova abordagem de testes para agentes deliberativos. [Title in English: JAT4BD: a new approach to testing deliberative agents]. MSc. Diss. Port. Presentation: 17/12/2015. 82 p. Advisor: Carlos José Pereira de Lucena.

Abstract: The growth and popularity of the Web has fueled the development of software-based network. The use of multi-agent systems (MAS) in this context is considered a promising approach has been applied in different areas such as security, or mission critical business scenarios, enhanced monitoring of environments and people, etc., which means analyzing the choices that this type of software can become crucial. However, the methodologies proposed so far by the Software Engineering Oriented Agents (AOSE) focused their efforts mainly on developing disciplined approach to analyze, design and implement an SMA and little attention has been given to how such systems can be tested. Furthermore, with regard to tests involving software agents, some issues related to the controllability and observability difficult the task of checking the behavior, such as: (i) the duration of the agent in its decision-making process; (ii) the fact of the agent's beliefs and goals are embedded in the agent itself, hampering the observation and control of behavior; (iii) problems associated with test coverage. In this research a novel approach for unit testing of agents written in BDI4JADE BDI based on the combination and arrangement of ideas supported by JAT Framework, a framework for testing agents written in JADE and fault model proposed by Zhang is displayed.

[15_MSc_amorim]
Franklin Anderson AMORIM. Mineração de itens frequentes em sequências de dados: uma implementação eficiente usando vetores de bits. [Title in English: Mining frequent itemsets in data streams: an efficient implementation using bit vectors]. MSc. Diss. Port. Presentation: 04/09/2015. 48 p. Advisor: Marco Antonio Casanova.

Abstract: The mining of frequent itemsets in data streams has several practical applications, such as user behavior analysis, software testing and market research. Nevertheless, the massive amount of data generated may pose an obstacle to processing then in real time and, consequently, in their analysis and decision making. Thus, improvements in the efficiency of the algorithms used for these purposes may bring great benefits for systems that depend on them. This thesis presents the MFI-TransSW+ algorithm, an optimized version of MFI-TransSW algorithm, which uses bit vectors to process data streams in real time. In addition, this thesis describes the implementation of a news articles recommendation system, called ClickRec, based on the MFI-TransSW+, to demonstrate the use of the new version of the algorithm. Finally, the thesis describes experiments with real data and presents results of performance and a comparison between the two algorithms in terms of performance and the hit rate of the ClickRec recommendation system.

[15_PhD_lima]
Guilherme Augusto Ferreira LIMA. A synchronous virtual machine for multimedia presentations. [Title in Portuguese: Uma máquina virtual síncrona para apresentações multimídia]. Ph. D. Thesis. Eng. Presentation: 01/12/2015. 134 p. Advisor: Luiz Fernando Gomes Soares.

Abstract: Current high-level multimedia languages are limited. Their limitation stems not from the lack of features but from the complexity caused by the excess of them and, more importantly, by their unstructured definition. Languages such as NCL, SMIL, and HTML define innumerable constructs to control the presentation of audiovisual data, but they fail to describe how these constructs relate to each other, especially in terms of behavior. There is no clear separation between basic and derived constructs, and no apparent principle of hierarchical build-up in their definition. Users may not need such principle, but it is indispensable for the people who define and implement these languages: it makes specifications and implementations manageable by reducing the language to a set of basic (primitive) concepts. In this thesis, a set of such basic concepts is proposed and taken as the language of a virtual machine for multimedia presentations. More precisely, a novel high-level multimedia language, called Smix (Synchronous Mixer), is presented and defined to serve as an appropriate abstraction layer for the definition and implementation of higher level multimedia languages. In defining Smix, that is, choosing a set of basic concepts, this work strives for minimalism but also aims at tackling major problems of current high-level multimedia languages, namely, the inadequate semantic models of their specifications and unsystematic approaches of their implementations. On the specification side, the use of a simple but expressive synchronous semantics, with a precise notion of time, is advocated. On the implementation side, a two-layered architecture that eases the mapping of specification concepts into digital signal processing primitives is proposed. The top layer (front end) is the realization of the semantics, and the bottom layer (back end) is structured as a multimedia digital signal processing dataflow.

[15_MSc_zampronio]
Guilherme Bezerra ZAMPRONIO. Um sistema de simulação 3D de evacuação de emergência em plataformas de petróleo. [Title in English: A 3D simulation system for emergency evacuation in offshore platforms]. M.Sc. Diss. Port. Presentation: 02/03/2015. 57 p. Advisor: Alberto Barbosa Raposo.

Abstract: An application for evacuation simulation using computational resources may help previewing situations, flows, conflicts, and behaviors that may only happen in a real danger situation. This kind of application enables the execution of several pre-defined scenarios at any time, without the expensive and complex allocation of real people. This dissertation proposes an emergency simulation system on oil platforms in 3D with real-time results using as architecture a game engine (Unity). The solution developed was tested in of real platforms models for comparison with times obtained in emergency simulations with people. System performance will be exposed, as well as future works.

[15_MSc_borges]
Heraldo Pimenta BORGES FILHO. Predição do comportamento do mercado financeiro utilizando notícias em Português. [Title in English: Stock market behavior prediction using financial news in Portuguese]. M.Sc. Diss. Port. Presentation: 29/08/2015. 51 p. Advisor: Ruy Luiz Milidiu.

Abstract: A set of financial theories, such as the efficient market hypothesis and the theory of random walk, says it is impossible to predict the future of the stock market based on currently available information. However, recent research has proven otherwise by finding a relationship between the content of a news and current behavior of an stock. Our goal is to develop and implement a prediction algorithm that uses financial news about joint-stock company to predict the stock's behavior on the stock exchange. We use an approach based on machine learning for the task of predicting the behavior of an stock in positions of up, down or neutral, using quantitative and qualitative information, such as financial. We evaluate our system on a dataset with six thousand news and our experiments indicate an accuracy of 68.57% for the task.

[15_MSc_gualandi]
Hugo Musso GUALANDI. Typing dynamic languages – a review. [Title in Portuguese: Phase unwrapping 2D via floresta geradora mínima com restrições de balanceamento]. M.Sc. Diss. Eng. Presentation: 08/09/2015. 89 p. Advisor: Roberto Ierusalimschy.

Abstract: Programming languages have traditionally been classified as either statically typed or dynamically typed, the latter often being known as scripting languages. Dynamically typed languages are very popular for writing smaller programs, a setting where ease of use and flexibility of the language are highly valued. However, with time, small scripts tend to evolve into large systems and the flexibility of the dynamic language may become a source of program defects. For these larger systems, static typing, which offers compile-time error detection, improved documentation and optimization opportunities, becomes more attractive. Since rewriting the whole system in a statically typed language is not ideal from a software engineering point of view, investigating ways of adding static types to existing dynamically typed programs has been a thriving research area. In this work, we present a historical overview of this research. We focus on general approaches that apply to multiple programming languages, such as the Type Hints of Common LISP, the Soft Typing of Fagan et al and the Gradual Typing of Siek et al, contrasting these different solutions from a modern perspective.

[15_MSc_herszterg]
Ian Hodara HERSZTERG. 2D phase unwrapping via minimum spanning forest with balance constraints. [Title in Portuguese: Phase unwrapping 2D via floresta geradora mínima com restrições de balanceamento]. M.Sc. Diss. Eng. Presentation: 31/08/2015. 89 p. Advisors: Thibault Vidal and Marcus Vinicius Soledade Poggi de Aragão (co-advisor).

Abstract: The development and application of techniques in coherent signal processing have greatly increased over the past several years. Synthetic aperture radar, acoustic imaging, magnetic resonance, X-Ray crystallography and seismic processing are just a few examples in which coherent processing is required, resulting in the need for accurate and efficient phase unwrapping methods. The phase unwrapping problem consists in recovering a continuous phase signal from an originally wrapped phase data between the [symbol] interval. This dissertation proposes a new model for the L0-norm 2D phase unwrapping problem, in which the singularities of the wrapped phase image are associated to a graph where the vertices have different polarities (+1 and -1). The objective is to find a minimum cost balanced spanning forest where the sum of polarities is equal to zero in each tree. A set of primal and dual heuristics, a branchand-cut algorithm and a hybrid metaheuristic method are proposed to address the problem, leading to an efficient approach for L0-norm 2DPU, previously viewed as highly desirable but intractable. A set of experimental results illustrates the effectiveness of the proposed algorithm, and its competitiveness with state-of-the-art algorithms.

[15_PhD_monteiro]
Ingrid Texeira MONTEIRO. Autoexpressão e engenharia semiótica do usuário-designer. [Title in English: User designer’s self-expression and semiotic engineering]. Ph.D. Thesis. Port. Presentation: 15/04/2015. 312 p. Advisor: Clarisse Sieckenius de Souza.

Abstract: This thesis presents research in the area the area of End-User Development (EUD). The first studies in EUD have emerged as an attempt to help end users achieve specific goals of personalization and customization of interfaces and systems, primarily for their own benefit. As needs evolve, end users have to know and often master more complex computing concepts and practices. In this context, there have been a growing number of initiatives to encourage, teach and support users in programming and thinking computationally. In general, much emphasis is given to problem solving, logical reasoning and other common computer scientists’ skills. However, supported by Semiotic Engineering, a semiotics-based theory that describes human-computer interaction as communication between designers and users, we believe that interactive computer systems are communication artifacts: that the person who creates the system sends various messages, with particular characteristics to the person who uses it. In this thesis, we present an extensive study in which end users, acting as designers, create computational artifacts for communication purposes. Research has shown that the participants took programming and other development activities not as end in themselves but as a means to build their messages. We discuss how the change in perspective (from problem-solving to communication) reveals a range of underexplored phenomena, such as self-expression of the designers and the pragmatics of interaction languages they build. Another contribution of this thesis is an extension to Semiotic Engineering, named EUME – End-User Semiotic Engineering, a new way to look at Semiotic Engineering, in the perspective of end users acting as designers.

 [15_PhD_barbosa]
Ivanildo BARBOSA. Avaliação do impacto de acidentes de trânsito no tráfego de vias urbanas a partir de valores de velocidade. [Title in English: Assessment of the impact of traffic accidents on the flow of urban roads based on speed values]. Ph.D. Thesis. Port. Presentation: 27/03/2015. 175 p. Advisor: Marco Antonio Casanova.

Abstract: A major concern in large cities is to minimize the effects of the increasing quantity of vehicles in circulation and, consequently, of the accidents that tend to occur more frequently. Due to the popularization and miniaturization of GPS receivers, the availability of large volumes of data about vehicle speed in urban roads and the large number of traffic-related messages published in social networks, it is now possible to collect enough input data to model traffic conditions based on the observed reduction in speed values. However, it is necessary to filter the data to minimize thematic, spatial and temporal uncertainties. This thesis proposes a methodology to assess the impact of traffic accidents by analyzing speed values. To achieve this goal, it also proposes auxiliary methodologies, aiming at: (1) processing GPS-tracked routes to compute speed statistics and estimate traffic in two-way streets, by performing direction analysis; (2) representing traffic behavior based on the observed speed values; (3) extracting and selecting accident-related data by mining Twitter posts for later identification of the likely effects on speed values. The main contributions of this thesis are: (1) the assessment of traffic conditions based on speed values, which are easier to acquire than data about traffic volume and concentration; (2) the use of posts from social networks, which provide timely access to traffic events; (3) the assessment of urban roads instead of freeways or roads, which require modeling intersections, traffic lights and pedestrian flow; and (4) a methodology designed to extract speed statistics from raw GPS data, which handles likely error sources related to both map matching process and temporal classification.

[15_MSc_escobar]
Jaisse Grela ESCOBAR. Uma ferramenta para o rastreamento de vídeos e imagens utilizando técnicas de esteganografia. [Title in Portuguese: A tool for tracking videos and images using steganography techniques]. MSc. Diss. Port. Presentation: 10/04/2015. 57 p. Advisor: Bruno Feijo.

Abstract: In the TV industry, leaks of film materials occur frequently when they are distributed among the members of the production team, causing great harm to the companies. In this paper, we propose a tool that allows detecting the source of the leak with a high degree of confidence, using techniques of adaptive steganography. An important requirement is that the information embedded in the video (or image) should resist to processing operations such as resizing and resolution changes. The idea is to use the “Speeded Up Robust Features” (SURF) algorithm, a well-known strategy for detection and description of images features, to detect the robust regions of the image and insert a small masked identification in them. The tool uses the "Haar - Discrete Wavelet Transform" in two dimensions and then modifies the image. This dissertation proposes promising initial directions for secure identification of the certificate of origin of digital images and videos.

[15_MSc_bastos]
João Antonio Dutra Marcondes BASTOS. Apoio à transferência de conhecimento de raciocínio computacional de linguagens de programação visuais para linguagens de programação textuais. [Title in English: Support for computational thinking knowledge transfer from visual programming languages to textual programming languages]. M.Sc. Diss. Port. Presentation: 13/04/2015. 76 p. Advisor: Clarisse Sieckenius de Souza.

Abstract: Producing technology has been an increasingly essential ability in modern society. The users are no longer simple consumers but actually, also, technology producers, using technology to express their ideas. In this context, the learning of the so-called "computational thinking" should be as important as learning basic disciplines such as reading, writing and arithmetic. As long as the student can develop this ability, he will be able to express himself or herself through the software. Many projects around the world have their own technologies and pedagogy to help the student develop such capacity. However, we know that in a context that is constantly evolving as is the case of informatics, we cannot allow the student to be attached to a single tool or means. Tools may become obsolete and students would lose their technology producer status. With this in mind, we designed a learning transfer model of "computational thinking", which will assist the designer in the creation of a technological artifact to help students and teachers learn a new programming language. The model, which is based on the Semiotic Engineering, is the main scientific contribution of this master's dissertation.

[15_PhD_ferreira]
Juliana Soares Jansen FERREIRA. Comunicação através de modelos no contexto do desenvolvimento de software.  [Title in English: Communication through models in the context of software development]. Ph.D. Thesis. Port. Presentation: 10/04/2015. 200 p. Advisor: Clarisse Sieckenius de Souza.

Abstract: Software development is a highly collaborative process where software construction is the common goal. It is supported at several stages by computer tools, including software modeling tools. Models are important artifacts of the software development process and constitute the focus of this research, which aims to investigate the ‘communicability’ of software models produced and consumed with the support of modeling tools. Software model communicability is the capacity that such artifacts have of carrying and effecting a communication process among people, or of being used as an instrument to perform a significant part of such process. Modeling tools have a direct impact in that communicability, since model’s producers and consumers interact with those tools throughout the software development process. During that interaction, software models, which are intellectual artifacts, are created, changed, evolved, transformed and shared by people involved in activities of specification, analysis, design and implementation of the software under development. Besides the influence of tools, software modeling also needs to take into consideration previously defined notations as premises for modeling activities. This research is an investigation on how tools and notations influence and support the intellectual process of production and consumption of software models. We have Semiotic Engineering as our guiding theory given the essence of it that is: a careful study of tools people interact with to build, use and publish models through which they coordinate teamwork. The use of models in the software development process is a phenomenon that includes several factors that cannot be isolated from each other. Therefore, we propose a Tool-Notation-People triplet (TNP triplet) as a means of articulation to characterize observed issues about models in the software development. Along with the TNP triplet, we introduce a method that combines semiotic and cognitive perspectives to evaluate software modeling tools, producing data about the emission of designer-user metacommunication, users in this case being software developers. We aim to track potential relations between the human-computer interaction experience of those involved in the software development process while creating/reading/editing models with: (a) the product (types of models) generated in the process; and (b) the interpretations that such models evoke when used effectively in everyday practical situations to ‘communicate and express ideas and understandings’. The interest of working with Semiotic Engineering in this research is twofold. First, as an ‘observation lens’, the theory offers many resources to investigate and understand the construction and use of computational artifacts, their meanings and roles in the communication process. Second, a better perspective about the complete process that results, ultimately, in the user experience during the interaction with the software is relevant for the theory’s own evolution. In other words, this research has produced further knowledge about the communication conditions and mutual understanding of those who, according to the theory, ‘communicate their intent and design principles through the interface’, a potentially valuable source of explanations about communication problems in HCI.

[15 MSc_fontoura]
Leonardo Lobo da Cunha da FONTOURA.  On the min distance superset problem. [Title in Portuguese: Sobre o problema de superset mínimo de distâncias. M.Sc. Diss. Eng. Presentation: 19/08/2015. 48 p. Advisors: Thibaut Vidal and Marcus Vinícius Soledade Poggi de Aragão.

Abstract: The Partial Digest Problem, also known as the Turnpike Problem, consists of building a set of points on the real line given their unlabeled pairwise distances. A variant of this problem, named Min Distance Superset Problem, deals with incomplete input in which distances may be missing. The goal is to find a minimal set of points on the real line such that the multiset of their pairwise distances is a superset of the input. The main contributions of this work are two different mathematical programming formulations for the Min Distance Superset Problem: a quadratic programming formulation and an integer programming formulation.We show how to apply direct computation methods for variable bounds on top of a Lagrangian relaxation of the quadratic formulation. We also introduce two approaches to solve the integer programming formulation, both based on binary searches on the cardinality of an optimal solution. One is based on a subset of decision variables, in an attempt to deal with a simpler feasibility problem, and the other is based on distributing available distances between possible points.

[15_MSc_benevides]
Leonardo de Paula Batista BENEVIDES. Oclusão de ambiente para renderização de linhas. [Title in English: Ambient occlusion for line rendering]. M.Sc. Diss. Port. Presentation: 25/09/2015. 66 p. Advisor: Waldemar Celes Filho.

Abstract: The three-dimensional understanding of dense line sets requires the use of more sophisticated lighting models. Ambient occlusion is a technique used to simulate realistically and efficiently, the indirect ambient lighting. This paper presents a new algorithm for rendering lines with ambient occlusion. The proposed algorithm is based on the voxelization of the scene and on the computation of occlusionin the hemisphere associated to each visible point. It is proposed an adaptation of the voxelization algorithm of 3D scenes made up of solids to the correct treatment of the scene formed by lines. Thus, a volumetric geometry description is created in a texture buffer. The hemisphere around every visible point is sampled by several points, and for each sample is generated a prism, which  occluded volume is calculated from the voxelization. By accumulating the results of each sample, the estimated ambient occlusion caused by the geometry at each point visible to the observer is computed. This strategy proved to be appropriate, resulting in high-quality images in real time for complex scenes.

[15_MSc_campagnolo]
Leonardo Quatrin CAMPAGNOLO.  Visualização volumétrica precisa baseada em integração numérica adaptativa. [Title in English: Accurate volume rendering based on adaptive numerical integration]. M.Sc. Diss. Port. Presentation: 19/08/2015. 76 p. Advisor: Waldemar Celes Filho.

Abstract: One of the main challenges in volume rendering algorithms is how to compute the Volume Rendering Integral accurately, while maintaining good performance. Commonly, numerical methods use equidistant samples to approximate the integral and do not include any error estimation strategy to control accuracy. As a solution, adaptive numerical methods can be used, because they can adapt the step size of the integration according to an estimated numerical error. On CPU, adaptive integration algorithms are usually implemented recursively. On GPU, however, it is desirable to eliminate recursive algorithms. In this work, an adaptive and iterative integration strategy is presented to evaluate the volume rendering integral for regular volumes, maintaining the control of the step size for both internal and external integrals. A set of computational experiments were made comparing both accuracy and efficiency against the Riemann summation with uniform step size. The proposed algorithm generates accurate results, with competitive performance. The comparisons were made using both CPU and GPU implementations.

[15_MSc_alves]
Luciana Brasil Sondermann ALVES. Um estudo sobre a captura de objetos em movimento com múltiplas câmeras RGB-D para efeitos especiais. [Title in English: A study about capture of objects in moving with mulpiple RGB-D]. M.Sc. Diss. Port. Presentation: 17/07/2015. 59 p. Advisor: Bruno Feijó.

Abstract: This dissertation is an investigation on the generation of visual effects (VFX) from the capture of moving objects as very dense point clouds using multiple low-cost RGB-D cameras. For this investigation, we use a commercial software for particle rendering and some modules already developed by the Department o VFX R&D of TV Glogo in partnership with ICAD/VisionLab laboratory for the simultaneous capture of multiple MS Kinect cameras.  In the proposed production pipeline, a server synchronizes the shooting of multiple cameras, unifies the clouds and generates a file in standard format (PRP or PLY). This file is then is used for volumetric particle rendering with added visual effects using the Krakatoa MX software for 3DS max. The goal is to shoot scenes in such a way that the film director can later define the final scenes with any camera path and adding visual effects.

[15_MSc_souza]
Luciano Sampaio Martins de SOUZA. Early vulnerability detection for supporting secure programming. [Title in Portuguese: Detecão da vulnerabilidade de segurança em tempo de programação com intuito de dar suporte à programação segura]. MSc. Diss. Eng. Presentation: 15/01/2015. 132 p. Advisor: Alessandro Fabrício Garcia.

Abstract: Secure programming is the practice of writing programs that are resistant to attacks by malicious people or programs. Programmers of secure software have to be continuously aware of security vulnerabilities when writing their program statements. They also ought to continuously perform actions for preventing or removing vulnerabilities from their programs. In order to support these activities, static analysis techniques have been devised to find vulnerabilities in the source code. However, most of these techniques are built to encourage vulnerability detection a posteriori, only when developers have already fully produced (and compiled) one or more modules of a program. Therefore, this approach, also known as late detection, does not support secure programming but rather encourages posterior security analysis. The lateness of vulnerability detection is also influenced by the high rate of false positives, yielded by pattern matching, the underlying mechanism used by existing static analysis techniques. The goal of this dissertation is twofold. First, we propose to perform continuous detection of security vulnerabilities while the developer is editing each program statement, also known as early detection. Early detection can leverage his knowledge on the context of the code being created, contrary to late detection when developers struggle to recall and fix the intricacies of the vulnerable code they produced from hours to weeks ago. Our continuous vulnerability detector is incorporated into the editor of an integrated software development environment. Second, we explore a technique originally created and commonly used for implementing optimizations on compilers, called data flow analysis, hereinafter referred as DFA. DFA has the ability to follow the path of an object until its origins or to paths where it had its content changed. DFA might be suitable for finding if an object has a vulnerable path. To this end, we have implemented a proof-of-concept Eclipse plugin for continuous vulnerability detection in Java programs. We also performed two empirical studies based on several industry-strength systems to evaluate if the code security can be improved through DFA and early vulnerability detection. Our studies confirmed that: (i) the use of data flow analysis significantly reduces the rate of false positives when compared to existing techniques, without being detrimental to the detector performance, and (ii) early detection improves the awareness among developers and encourages programmers to fix security vulnerabilities promptly.

[15_MSc_souza]
Luiz Gustavo de SOUZA.  Estendendo a MoLIC para apoiar o design de sistemas colaborativos. [Title in English: Extending MoLIC to support the design of collaborative systems]. M.Sc. Diss. Port. Presentation: 08/09/2015. 126 p. Advisor: Simone Diniz Junqueira Barbosa.

Abstract: The field of Computer-Supported Cooperative Work focuses on the understanding of collaborative systems and methodologies for the design and development of such systems. The 3C Collaboration Model divides the collaboration process into communication, coordination and cooperation. Regarding Human-Computer Interaction, different interaction models aim to support designers in the user and system interaction design. Semiotic Engineering views the design and use of technology as communication processes. It contributes with interaction design models, such as MoLIC, a language that supports the design of the interaction between the user and the designer’s deputy (the user interface). The original MoLIC language provides no support for collaborative systems design, raising the need for a study that considers these limitations, addressing questions in order to understand the interaction design of collaborative systems based on the 3C Model using MoLIC. The present work presents a review on MoLIC focusing collaborative systems, presenting the extension MoLICC, whose effectiveness for collaborative systems design we evaluated by conducting an empirical study with users and analyzing the language using the Cognitive Dimensions of Notations framework.

[15_PhD_afonso]
Luiz Marques AFONSO.  Communicative dimensions of application programming interfaces (APIs). [Title in Portuguese: Dimensôes Comunicativas de Interfaces de Programação (APIs)]. Ph.D. Thesis. Eng. Presentation: 06/04/2015. 153 p. Advisor: Clarisse Sieckenius de Souza.

Abstract: Application programming interfaces (APIs) have a central role in software development, as programmers have to deal with a number of routines and services that range from operating system libraries to large application frameworks. In order to effectively use APIs, programmers should have good comprehension of these software artifacts, making sense of the underlying abstractions and concepts by developing an interpretation that is compatible with the designer's intent. Due to the complexity of today's systems and programming environments, learning and using an API properly can be non-trivial task to many programmers. Traditionally, studies on API design have been developed from a usability standpoint. These studies have provided evidence that bad APIs may a_ect programmer's productivity and software quality, offering valuable insights to improve the design of new and existing APIs. This thesis proposes a novel approach to investigate and discuss API design, based on a communication perspective under the theoretical guidance of Semiotic Engineering. From this perspective, an API can be viewed as a communication process that takes place between designer and programmer, in which the former encodes a message to the latter about how to communicate back with the system and use the artifact's features, according to its design vision. This approach provides an account of API design space that highlights the pragmatic and cognitive aspects of human communication mediated by this type of software artifact. By means of the collection and qualitative analysis of empirical data from bug repositories and other sources, this research work contributes to a deeper comprehension of the subject, providing an epistemic framework that intends to support the analysis, discussion and evaluation of API design.

[15_MSc_costa]
Marcelo Arza Lobo da COSTA.  Análise de duty-cycling para economia de energia na disseminação decódigo em rede de sensores. [Title in English: Analysis of duty-cycling for saving energy in code dissemination over sensor networks]. M.Sc. Diss. Port. Presentation: 08/09/2015. 49 p. Advisor: Noemi de La Roque Rodriguez.

Abstract: One of the key challenges in wireless sensor networks (WSN) is to save energy at motes. One method to save battery is radio duty cycling (DC), which keeps the radio turned off in most of the time and turns the radio on for a short time to verify if any there are any messages. DC is frequently used in monitoring applications where only one message is transmitted after the mote reads its sensor. Usually the mote reads its sensor only once every few minutes, so few unicast messages are transmitted in the network per time unit. This work analyzes the use of the DC method in code dissemination. In this context, multiple broadcast messages are transmitted in a short time. We examined two specific dissemination algorithms, one of them proposed for a virtual machine environment, in which the disseminated code is a small script, and a second one originally proposed for disseminating the code of an entire application, typically much larger than a script. The objective of this study is to evaluate the impact of DC on latency and how much energy was saved when compared to leaving the radio on all the time, which is how both algorithms work in their original form.

[15_MSc_nascimento]
Marcelo de Mattos NASCIMENTO. Utilizando reconstrução 3D densa para odometria visual baseada em técnicas de structure from motion. [Title in English: Using dense 3D reconstruction for visual odometry based on structure from motion techniques]. M.Sc. Diss. Port. Presentation: 17/09/2015.  59 p. Advisors: Alberto Barbosa Raposo.

Abstract: Aim of intense research in the field computational vision, dense 3D reconstruction achieves an important landmark with first methods running in real time with millimetric precision, using RGBD cameras and GPUs. However these methods are not suitable for low computational resources. Having low computational resources as requirement, the goal of this work is to show a method of visual odometry using regular cameras, without using a GPU. The proposed method is based on technics of sparse Structure From Motion (SFM), using data provided by dense 3D reconstruction. Visual odometry is the process of estimating the position and orientation of an agent (a robot, for instance), based on images. This dissertation compares the proposed method with the odometry calculated by Kinect Fusion. Results of this research are applicable in augmented reality. Odometry provided by this work can be used to model a camera and the data from dense 3D reconstruction, can be used to handle occlusion between virtual and real objects.

[15_PhD_alencar]
Marcus Franco Costa de ALENCAR. A Context-aware model for decision-making within multi-expertise tactical planning in a major incident. [Title in Portuguese: Modelo ciente de contexto para tomada de decisão no planejamento tático multi-especialidade em um grande incidente]. Ph.D. Thesis. Eng. Presentation: 27/02/2015.  112 p. Advisors: Alberto Barbosa Raposo and Simone Diniz Junqueira Barbosa.

Abstract: Activities that involve very complex and unpredictable problems are still relying on human decision-making and manual tools like paper forms. This is still the case for decision-making within tactical planning in a major incident, where the collective knowledge of multi-expertise specialists is essential to make urgent and effective decisions. Specialists tend to reject tools that jeopardize the decision agility required, as it can put human lives at risk and cause major damages to the environment and property. Communications and management are challenges, as these specialists have each their own “expert language”. This thesis proposes a planning model that keeps decisions with specialists and allows them to use their own expressions for tactical planning, but incorporates expressions’ context traceability, authority control over these expressions, and expressions storage for reuse. These features were implemented in a Web tool that aims at empowering decision-making specialists, but tries to preserve the attributes of paper forms. The Web tool was evaluated in oil & gas industry emergency scenarios, and evaluation results indicated that the model’s approach enables important improvements in tactical planning communications and management within this context.

[15_MSc_lami]
Milena Ossorio LAMI. Um assistente dirigido por modelos para auxílio ao desenvolvimento de aplicações WWW. [Title in English: A Model-driven wizard to aid in developing Web applications]. M.Sc. Diss. Port. Presentation: 16/03/2015.  96 p. Advisor: Daniel Schwabe.

Abstract: Web applications can be seen as examples of hypermedia applications. Developing such applications is a complex endeavor, even when using design methods. There are model-driven methods aimed at helping the designer, but they still require a steep learning curve for those unfamiliar with the models. This work addresses this problem through a model-driven wizard that helps the designer through the use of examples and concrete data-driven interfaces. This wizard uses direct manipulation techniques to help easing the designer’s tasks.

[15_MSc_moreira]
Nara Torres MOREIRA. A MIP-based approach to solve a real-world school timetabling problem. [Title in Portuguese: Uma abordagem baseada em programação inteira mista para resolver um problema do mundo real de geração de grades horárias escolares]. M.Sc. Diss. Port. Presentation: 21/05/2015.  136 p. Advisor: Marcus Vinicius Soledade Poggi de Aragão.

Abstract: Timetabling problems look to schedule meetings in order to satisfy a set of demands, while respecting additional constraints. In a good solution the resulting timetables are acceptable to all people and resources involved. In school timetabling, a given number of lectures, involving students, teachers and classrooms, need to be scheduled over the week, while having to satisfy operational, institutional, pedagogical and personal restrictions. The di_culty of the problem has driven many researchers to work on solving approaches for it since the early 1960's. Finding an actual solution to a real world scenario implies satisfying many quality requirements and not ignoring the political issues, which turns the classical problem much more intricate. This work describes an approach based on mixed integer programming (MIP) developed for solving a real-world school timetabling problem and discusses ideas and issues faced during solution deployment phase for some Brazilian schools. In contrast to other works on school timetabling, teaching sta_ sharing between distinct school units are considered. Computational experiments were performed for scenarios whose number of school units varies from 2 to 15, number of teachers varies from 35 to 471 and number of classes varies from 16 to 295. Di_erent strategies were combined aiming at converging to good solutions. Finally, results are evaluated and the best approaches are highlighted.

[15_MSc_moreira]
Nathalia Moraes do NASCIMENTO. FloT: an agent-based framework for selfadaptive and self-organizing Internet of Things applications. [Title in Portuguese: FIoT: um framework baseado em agentes para aplicações auto-organizáveis e autoadaptativas de Internet das Coisas]. M.Sc. Diss. Eng. Presentation: 31/08/2015.  102 p. Advisor: Carlos José Pereira de Lucena.

Abstract: The agreed fact about the Internet of Things (IoT) is that, within the coming years, billions of resources, such as cars, clothes and foods will be connected to the Internet. However, several challenging issues need to be addressed before the IoT vision becomes a reality. Some open problems are related to the need of building self-organizing and self-adaptive IoT systems. To create IoT applications with these features, this work presents a Framework for Internet of  Things (FIoT). Our approach is based on concepts from Multi-Agent Systems (MAS) and Machine Learning Techniques, such as a neural network and evolutionary algorithms. An agent could have characteristics, such as autonomy and social ability, which makes MAS suitable for systems requiring self-organization (SO). Neural networks and algorithms of evolution have been commonly used in robotic studies to provide embodied agents (as robots and sensors)with autonomy and adaptive capabilities. To illustrate the use of FIoT, we derived two different instances from IoT applications: (i) Quanti ed Things and (ii) Smart Cities. We show how flexible points of our framework are instantiated to generate an application.

[15_MSc_musa]
Pablo Martins MUSA. Profiling memory in Lua. [Title in Portuguese: Analisando o uso de Memória em Lua]. M.Sc. Diss. Port. Presentation: 19/06/2015. 89 p. Advisor: Roberto Ierusalimschy.

Abstract: Memory bloat is a software problem that happens when the memory consumption of a program exceeds the programmer's expectations. In many cases, memory bloat hurts performance or even crashes applications. Detecting and fixing memory bloat problems is a difficult task for programmers and, thus, they usually need tools to identify and fix these problems. The past two decades produced an extensive research and many tools to help programmers tackle memory bloat, including memory profilers. Although memory profilers have been largely studied in the last years, there is a gap regarding scripting languages. In this thesis, we study memory profilers in scripting languages. First, we propose a classification in which we divide memory profilers in manual and automatic, based on how the programmer uses the memory profiler. Then, after reviewing memory profilers available in three different scripting languages, we experiment some of the studied techniques by implementing two automatic memory profilers to help Lua programmers deal with memory bloat. Finally, we evaluate our tools regarding how easy it is to incorporate them to a program, how useful their reports are to understand an unknown program and track memory bloats, and how much overhead they impose.

[15_PhD_pampanelli]
Patrícia Cordeiro Pereira PAMPANELLI. Suavização de dados de amplitude através de difusão anisotrópica com preservação de feições sísmicas. [Title in English: Seismic amplitude smoothing by anisotropic diffusion preserving structural features]. Ph.D. Thesis. Port. Presentation: 13/08/2015. 95 p. Advisor: Marcelo Gattass.

Abstract: Seismic interpretation can be viewed as a set of methodologies to enhance the understanding of the structural and stratigraphic model of a given region. During this process, the interpreter analyzes the seismic imaging seeking to identify geological structures such as faults, horizons and channels, among  others. Given the low signal to noise ratio, the algorithms that support the interpretation require a pre-processing stage where the noise is reduced. This thesis proposes a new filtering method based on the anisotropic diffusion of the amplitude field. The formulation of the diffusion process proposed here uses seismic attributes to identify horizons and faults that are preserved in the diffusion process. The proposed method implemented in this thesis also presents results applied to real and synthetic data. Based on these results, we present an analysis of the influence of the proposed method in correlation measurements over horizons previously tracked. Finally the thesis presents some conclusions and suggestions for future work.

[15_MSc_diniz]
Pedro Henrique Fonseca da Silva DINIZ. A spatio-temporal model for average speed prediction on roads. [Title in Portuguese: Um modelo espaço-temporal para a previsão de velocidade média em estradas]. M.Sc. Diss. Eng. Presentation: 21/08/2015.  75 p. Advisor: Hélio Côrtes Vieira Lopes.

Abstract: Many factors may influence a vehicle in speed in a road, but two of them are usually observed by many drivers: its location and the time of the day.  To obtain a model that returns the average speed as a function of position and time is still a challenging task. The application of this models can be in different scenarios, such as: estimated time of arrival, shortest route paths, traffic prediction, and accident detection, just to cite a few.  This study proposes a prediction model based on spatial-temporal partitions and mean/instantaneous speeds collected from GPS data.  The main advantage of the proposed  model is that it is very simple to compute. Moreover, experimental results obtained from fuel delivery trucks, along the whole year of 2013 in Brazil, indicate that most of the observations can be predicted using this model within an acceptable error tolerance. 

[15_PhD_riverasalas]
Percy Enrique RIVERA SALAS. OLAP2Datacube: an on-demand transformation framework from OLAP to RDF data cubes. [Title in Portuguese: OLAP2Datacube: Um framework para transformações em tempo de execução de OLAP para cubos de dados em RDF]. Ph.D. Thesis. Eng. Presentation: 18/09/2015. 95 p. Advisor: Marco Antonio Casanova.

Abstract: Statistical data is one of the most important sources of information, relevant to a large number of stakeholders in the governmental, scientific and business domains alike. A statistical data set comprises a collection of observations made at some points across a logical space and is often organized as what is called a data cube. The proper definition of the data cubes, especially of their dimensions, helps processing the observations and, more importantly, helps combining observations from different data cubes. In this context, the Linked Data principles can be profitably applied to the definition of data cubes, in the sense that the principles offer a strategy to provide the missing semantics of the dimensions, including their values. In this thesis we describe the process and the implementation of a mediation architecture, called OLAP2DataCube On Demand, which helps describe and consume statistical data, exposed as RDF triples, but stored in relational databases. The tool features a catalogue of Linked Data Cube descriptions, created according to the Linked Data principles. The catalogue has a standardized description for each data cube actually stored in each statistical (relational) database known to the tool. The tool offers an interface to browse the linked data cube descriptions and to export the data cubes as RDF triples, generated on demand from the underlying data sources. We also discuss the implementation of sophisticated metadata search operations, OLAP data cube operations, such as slice and dice, and data cube mashup operations that create new cubes by combining other cubes.

[15_PhD_pereira]
Rafael Silva PEREIRA. A cloud based real-time collaborative filtering architecture for short-lived video recommendations. [Title in Portuguese: Uma arquitetura de filtragem colaborativa em tempo real baseada em nuvem para recomendação de vídeos efêmeros]. Ph.D. Thesis. Eng. Presentation: 11/12/2015. 86 p. Advisor: Hélio Cortes Vieira Lopes.

Abstract: This dissertation argues that the combination of collaborative filtering techniques, particularly for item-item recommendations, with emergent cloud computing technology can drastically improve algorithm efficiency, particularly in situations where the number of items and users scales up to several million objects. It introduces a real-time item-item recommendation architecture, which rationalizes the use of resources by exploring on-demand computing. The proposed architecture provides a real-time solution for computing online item similarity, without having to resort to either model simplification or the use of input data sampling. This dissertation also presents a new adaptive model for implicit user feedback for short videos, and describes how this architecture was used in a  large scale implementation of a video recommendation system in use by the largest media group in Latin America, presenting results from a real life case study to show that it is possible to greatly reduce recommendation times (and overall financial costs) by using dynamic resource provisioning in the Cloud. It discusses the implementation in detail, in particular the design of cloud based features. Finally, it also presents potential research opportunities that arise from this paradigm shift.

[15_MSc_diniz]
Rafael DINIZ. O Perfil NCL-DR e o middleware Ginga para receptores do Sistema Brasileiro de Rádio Digital. [Title in English: The NCL-DR profile and the Ginga middleware for the Brazilian Digital Radio System]. M.Sc. Diss. Port. Presentation: 10/07/2015.  135 p. Advisor: Luiz Fernando Gomes Soares.

Abstract: In 2010, the Minister of Communications instituted the Brazilian Digital Radio System (SBRD), however a reference model for the system has not yet been set. This text presents an analysis of the terrestrial radio broadcasting in Brazil and presents some requirements for an interactive digital radio. These requirements are then used to guide the research. The relevance and importance of the use of NCL and Ginga in the Brazilian digital radio as in Brazilian Digital TV System (SBTVD), are discussed, and the definition of how the transport of NCL applications should be done in the two digital radio systems being considered for adoption by the country is performed. A new profile of NCL for use in digital radio is defined. This new profile was named Digital Radio Profile, or just DR. Ginga is also defined for use in digital radio receivers, and new media players and features adapted to the digital radio context are introduced. PUC-Rio’s implementation of the Ginga middleware is presented with the necessary modifications for digital radio use. In addition, an environment to validate the interactivity in digital radio with the Ginga was assembled and tests exercising the entire digital radio chain of transmission and reception with embedded NCL applications were performed. The definitions and conclusions that resulted from research activities are expected to contribute to the definition of a Brazilian Digital Radio System that is powerful in communicational terms and aggregates the most relevant technologies for the medium in the digital age.

[15_MSc_oliveira]
Rafael Pereira OLIVEIRA. Sintonia fina baseada em ontologia: o caso de visões materializadas. [Title in English: Ontology-based database tuning: the case of materialized views]. M.Sc. Diss. Port. Presentation: 25/03/2015.  101 p. Advisor: Sérgio Lifschitz.

Abstract: The Outer-tuning framework may be used to support automatic (or not) database tuning, particularity index. It is an approach that offers transparency about the available alternatives to feasible tuning scenarios, making it possible to combine either independent strategies or allow discussion of justifications for actions performed in order to obtain better performances. Using a specific ontology for fine tuning relational databases, we add semantics to the process with the understanding of the concepts involved and generate (semi)automatic new tuning actions, which can be inferred from existing practices or new rules and concepts that arise in the future. This research presents as an initial contribution the actual design and implementation of the Outer-tuning framework through the formalization of a software architecture that meets the specified functional requirements. This work also contributes with the extension of the domain ontology and the inclusion of new heuristics to a task ontology, in order to accomplish fine tuning solutions with the use of materialized views. Thus, it becomes possible to propose the use of tuning heuristics for indexes as well as for materialized views.

[15_PhD_brandao]
Rafael Rossi de Mello BRANDÃO. A Capture & Access technology to support documentation and tracking of qualitative research applied to HCI. [Title in Portugues: Uma tecnologia de Captura & Acesso para suportar documentação e rastreamento de pesquisa qualitativa aplicada a IHC]. Ph.D. Thesis. Eng. Presentation: 08/04/2015. 153 p. Advisor: Clarisse Sieckenius de Souza.

Abstract: Tracking and exposure of qualitative methodology procedures is a problem observed in the scientific community. The traditional form of research publication makes it impractical to provide in detail all the decisions and evidences considered in the course of a qualitative research. To overcome this problem we propose an approach to structure all the procedures undertaken into hypermedia documents with analyses and validations, allowing its representation in a theoretical Capture & Access (C&A) model. This model enables the outlining of the research inquiry, providing semantics to allow relationship between key elements in a qualitative methodology. We discuss about five qualitative studies that guided the reasoning about the proposed model, pondering on how to register adequately the activities performed in HCI evaluations consolidating the collected data in documents used in posterior analysis sessions. Additionally, we present a proof of concept through an implementation using the C&A software infrastructure offered by the CAS Project. This infrastructure supports the recording of empirical data (text, images, audio, video, and slides), data post-processing and the generation of multimedia documents. It is possible to use tags for temporal annotation, create contexts to link data and retrieve other relevant information from the captured investigation processes.

[15_PhD_gomes]
Raphael do Vale Amaral GOMES. Crawling the linked data cloud. [Title in Portuguese: Coleta de dados interligados]. Ph.D. Thesis. Eng. Presentation: 12/05/2015. 118 p. Advisor: Marco Antonio Casanova.

Abstract: The linked data best practices recommend to publish a new tripleset using well-known ontologies and to interlink the new tripleset with other triplesets. However, both are difficult tasks. This thesis describes frameworks for metadata crawlers that help selecting the ontologies and triplesets to be used, respectively, in the publication and the interlinking processes. Briefly, the publisher of a new tripleset first selects a set of terms that describe the application domain of interest. Then, he submits the set of terms to a metadata crawler, constructed using one of the frameworks described in the thesis, that searches for triplesets which vocabularies include terms direct or transitively related to those in the initial set of terms. The crawler returns a list of ontologies that are used for publishing the new tripleset, as well as a list of triplesets with which the new tripleset can be interlinked. Hence, the crawler focuses on specific metadata properties, including subclass of, and returns only metadata, which justifies the classification “metadata focused crawler.

[15_MSc_rottgen]
Raphael Alexander ROTTGEN. Institutional ownership as a predictor of future security returns. [Title in English: Uso de dados das carteiras de investidores institucionais na predição de retornos de ações]. M.Sc. Diss. Eng. Presentation: 16/06/2015. 96 p. Advisor: Eduardo Sany Laber.

Abstract: Data on institutional ownership of securities is nowadays publicly available in a number of jurisdictions and can thus be used in models for the prediction of security returns. A number of recently launched investment products explicitly use such institutional ownership data in security selection. The purpose of the current study is to apply statistical learning algorithms to institutional ownership data from the United States, in order to evaluate the predictive validity of features based on such institutional ownership data with regard to future security returns. Our analysis identified that a support vector machine managed to classify securities, with regard to their four-quarter forward returns, into three bins with significantly higher accuracy than pure chance would predict. Even higher accuracy was achieved when "predicting" realized, i.e. past, fourquarter returns.

[15_PhD_berardi]
Rita Cristina Galarraga BERARDI. Design rationale in the triplification of relational databases. [Title in Portuguese: Design rationale na triplificação de bancos de dados relacionais]. Ph.D. Thesis. Eng. Presentation: 04/02/2015. 100 p. Advisor: Marco Antonio Casanova.

Abstract: One of the most popular strategies to publish structured data on the Web is to expose relational databases (RDB) in the RDF format. This process is called in RDB-to-RDF or triplification. Furthermore, the Linked Data principles offer useful guidelines for this process. Broadly stated, there are two main approaches to map relational databases into RDF: (1) the direct mapping approach, where the database schema is directly mapped to an RDF schema; and (2) the customized mapping approach, where the RDF schema may significantly differ from the original database schema. In both approaches, there are challenges related to the publication and to the consumption of the published data. This thesis proposes the capture of design rationale as a valuable source of information to minimize the challenges in RDB-to-RDF processes. Essentially, the capture of design rationale increases the awareness about the actions taken over the relational database to map it as an RDF dataset. The main contributions of this thesis are: (1) a design rationale (DR) model adequate to RDB-to-RDF processes, independently of the approach (direct or customized) followed; (2) the integration of a DR model in an RDB-to-RDF direct mapping process and in an RDB-to-RDF customized mapping process using the R2RML language; (3) the use of the DR captured to improve the recommendations for vocabularies to reuse.

[15_PhD_silva]
Rodrigo Marques Almeida da SILVA. Um método otimizado de renderização fotorealista com distribuição estatística e seleção automática de técnicas. [Title in Portuguese: An optimized photorealistic rendering method with statistic distribution and automatic rendering technique selection]. Ph.D. Thesis. Eng. Presentation: 09/04/2015. 180 p. Advisor: Bruno Feijó.

Abstract: The photorealistic rendering process for cinema and TV increasingly demands processing power, requiring fast parallel algorithms and effective task distribution systems. However, the processes currently used by the academia and by the industry still consume several days to evaluate an animation in super-resolution (typically 8K), what makes difficult the improvement of artistic quality and limits the number of experiments with scene parameters. In this work, we focus on the optimization of three processes involved in photorealistic rendering, reducing the total time of rendering substantially. Firstly, we optimize the local rendering, in which the system works to render a set of pixels optimally, taking advantage of the available hardware resources and using previous rendering data. Secondly, we optimize the management process, which is changed not only to distribute frames but also to analyze all the rendering infrastructure, optimizing the distribution process and allowing the establishment of goals as time and cost. Furthermore, the management model is expanded to the cloud, using the cloud as a processing overflow. Thirdly, we propose a new optimized process to evaluate the rendering task collaboratively, where each node communicates partial results to other nodes, allowing the optimization of the rendering process in all nodes. Altogether, this thesis is an innovative proposal to improve the whole process of high-performance rendering, removing waste of resources and reducing rework.

[15_MSc_magalhaes]
Rômulo de Carvalho MAGALHÃES. Operations over lightweight ontologies. [Title in Portuguese: Operações sobre ontologias leves]. MSc. Diss. Eng. Presentation: 30/01/2015. 106 p. Advisor: Marco Antonio Casanova.

Abstract: This work addresses ontology design problems by treating ontologies as theories and by defining a set of operations that map ontologies into ontologies, including their constraints. The work first summarizes the base knowledge needed to define the class of ontologies used and proposes four operations to manipulate them. It then shows how the operations work and how they may help design new ontologies. The core of this work is describing the implementation of the operations over a Protégé plug-in, detailing the architecture and including case-use examples.

[15_PhD_araujo]
Thiago Pinheiro de ARAÚJO. Using runtime information and maintenance knowledge to assist failure diagnosis, detection and recovery. [Title in Portuguese: Utilizando informações da execução do sistema e conhecimentos de manutenção para auxiliar o diagnóstico, detecção e recuperação de falhas]. Ph.D. Thesis. Eng. Presentation: 07/10/2015. 192 p. Advisor: Arnd von Staa.

Abstract: Even software systems developed with strict quality control may expect failures during their lifetime. When a failure is observed in a production environment the maintainer is responsible for diagnosing the cause and eventually removing it. However, considering a critical service this might demand too long a time to complete, hence, if possible, the failure signature should be identified in order to generate a recovery mechanism to automatically detect and handle future occurrences until a proper correction can be made. In this thesis, recovery consists of restoring a correct context allowing dependable execution, even if the causing fault is still unknown. To be effective, the tasks of diagnosing and recovery implementation require detailed information about the failed execution. Failures that occur during the test phase run in a controlled environment, allow adding specific code instrumentation and usually can be replicated, making it easier to study the unexpected behavior. However, failures that occur in the production environment are limited to the information present in the first occurrence of the failure. But run time failures are obviously unexpected, hence run time data must be gathered systematically to allow detecting, diagnosing with the purpose of recovering, and eventually diagnosing with the purpose of removing the causing fault. Thus there is a balance between the detail of information inserted as instrumentation and the system performance: standard logging techniques usually present low impact on performance, but carry insufficient information about the execution; while tracing techniques can record precise and detailed information, however are impracticable for a production environment. This thesis proposes a novel hybrid approach for recording and extracting system’s runtime information. The solution is based on event logs, where events are enriched with contextual properties about the current state of the execution at the moment the event is recorded. Using these enriched log events a diagnosis technique and a tool have been developed to allow event filtering based on the maintainer’s perspective of interest. Furthermore, an approach using these enriched events has been developed that allows detecting and diagnosing failures aiming at recovery. The proposed solutions were evaluated through measurements and studies conducted using deployed systems, based on failures that actually occurred while using the software in a production context.

 [15_PhD_silva]
Thuener Armando da SILVA. Optimization under uncertainty for asset allocation. [Title in Portuguese: Otimização sob incerteza para alocação de ativos]. Ph.D. Thesis. Presentation: 06/04/2015. Eng. 99 p. Advisor: Marcus Vinicius Soledade Poggi de Aragão and Michel Davi Valladão.

Abstract: Asset allocation is one of the most important financial decisions made by investors. However, human decisions are not fully rational, and people make several systematic mistakes due to overconfidence, irrational loss aversion and misuse of information, among others. In this thesis, we developed two distinct methodologies to tackle this problem. The first approach has a more qualitative view, trying to map the investor's vision of the market. It tries to mitigate irrationality in decision-making by making it easier for an investor to demonstrate his/her preferences for specific assets. This first research uses the Black-Litterman model to construct portfolios. Black and Litterman developed a method for portfolio optimization as an improvement over the Markowitz model. They suggested the construction of views to represent an investor's opinion about future stocks' returns. However, constructing these views has proven difficult, as it requires the investor to quantify several subjective parameters. This work investigates a new way of creating these views by using Verbal Decision Analysis. The second research focuses on quantitative methods to solve the multistage asset allocation problem. More specifically, it modifies the Stochastic Dynamic Dual Programming (SDDP) method to consider real asset allocation models. Although SDDP is a consolidated solution technique for large-scale problems, it is not suitable for asset allocation problems due to the temporal dependence of returns. Indeed, SDDP assumes a stagewise independence of the random process assuring a unique cost-to-go function for each time stage. For the asset allocation problem, time dependency is typically nonlinear and on the left-hand side, which makes traditional SDDP inapplicable. This thesis proposes an SDDP variation to solve real asset allocation problems for multiple periods, by modeling time dependence as a Hidden Markov Model with concealed discrete states. Both approaches were tested in real data and empirically analyzed. The contributions of this thesis are the methodology to simplify portfolio construction and the methods to solve real multistage stochastic asset allocation problems.

[15_MSc_faria]
Waldecir Vicente FARIA. D-Engine: a framework for the random execution of plans in agent-based models. [Title in Portuguese: D-Engine: um framework para a execução aleatória de planos em modelos baseados em agentes]. M.SC. Diss. Eng. Presentation: 10/07/2015. 72 p. Advisor: Hélio Cortes Vieira Lopes.

Abstract: An important question in agent-based systems is how to execute some planned action in a random way. The answer for this question is fundamental to keep the user's interest in some product, not just because it makes the experience less repetitive but also because it makes the product more realistic. This kind of action execution can be mainly applied on simulators, serious and entertainment games based on agent models. Sometimes the randomness can be reached by just generating random numbers. However, when creating a more complex product, it is recommended to use some statistical or stochastic knowledge to not ruin the product's consumption experience. In this work we try to give support to the creation of dynamic and interactive animation and story using an arbitrary model based on agents. Inspired on stochastic methods, we propose a new framework called D-Engine, which is able to create a random, but with a well-known expected behavior, set of timestamps describing the execution of an action in a discrete way following some specificate. While these timestamps allow us to animate a story, an action or a scene, the mathematical results generated with our framework can be used to aid other applications such as result forecasting, nondeterministic planning, interactive media and storytelling. In this work we also present how to implement two different applications using our framework: a duel scenario and an interactive online auction website.

[15_MSc_oizumi]
Willian Nalepa OIZUMI. Synthesis of code anomalies: revealing design problems in the source code. [Title in Portuguese: D-Engine: um framework para a execução aleatória de planos em modelos baseados em agentes]. M.SC. Diss. Eng. Presentation: 02/09/2015. 103 p. Advisor: Alessandro Fabricio Garcia.

Abstract: Design problems affect almost all software projects and make their maintenance expensive and impeditive. As design documents are rarely available, programmers often need to identify design problems from the source code. However, the identification of design problems is not a trivial task for several reasons. For instance, the reification of a design problem tends to be scattered through several anomalous code elements in the implementation. Unfortunately, previous work has wrongly assumed that each single code anomaly { popularly known as code smell { can be used as an accurate indicator of a design problem. There is growing empirical evidence showing that several types of design problems are often related to a set of inter-related code anomalies, the so-called code-anomaly agglomerations, rather than individual anomalies only. In this context, this dissertation proposes a new technique for the synthesis of code-anomaly agglomerations. The technique is intended to: (i) search for varied forms of agglomeration in a program, and (ii) summarize different types of information about each agglomeration. The evaluation of the synthesis technique was based on the analysis of several industry-strength software projects and a controlled experiment with professional programmers. Both studies suggest the use of the synthesis technique helped programmers to identify more relevant design problems than the use of conventional techniques