Theses and Dissertations

2011

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 2011.  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: 18/JUNE/2012
 

INDEX


[11_MSc_branco]
Adriano Francisco BRANCO. Um modelo de programação para RSSF com suporte à reconfiguração dinâmica de aplicações. [Title in English: A WSN programming model with a dynamic reconfiguration support]. M.Sc. Diss. Port. Presentation: 06/04/11.  98 p. Advisors: Noemi de La Rocque Rodriguez and Silvana Rosseto (UFF).

Abstract: Some basic characteristics of wireless sensor networks (WSN) make application creation and reconfiguration dificult tasks. A programming model is presented to simplify these tasks. This model is based on a set of parametrized components and on a Finite State Machine, and allows the remote configuration of different applications over the same set of installed components. We describe some tests to evaluate its impact on the development process, and the ease of applying modifications to a running application. We also measure the additional impact of remote configuration on network activity.


[11_MSc_silva]
Aleksander Medella Campos da SILVA. Uma biblioteca de componentes de software para simuladores de radar. [Title in English: A software component library for radar simulators]. M.Sc. Diss. Port. Presentation: 24/08/11. 95 p. Advisors: Renato Fontoura de Gusmão Cerqueira and Edward Hermann Haeusler.

Abstract: Nowadays Radar systems are becoming complex systems. Many radars are constituted by an array of elements, where a minimum of coordination among the elements functionality is need. The task to build a prototype in order to validate a radar design is an expensive one. On the other hand, there are many common features shared among diferent radar architectures. This dissertation describes the architecture of a simulator that is able to represent most of the radar systems designed according the basic principles of radars. The simulator is designed following a exible component based architecture and five quite diferent kind of radars are described and simulated using the presented architecture. Real scenarios are taken into account in the validation of the simulations.


[11_MSc_valeriano]
Allan Alves VALERIANO. Um mecanismo de seleção de componentes para o middleware Kaluana usando a noção de contratos de reconfiguração. [Title in English: A mechanism of component selection with the notion of reconfiguration contracts]. M.Sc. Diss. Port. Presentation: 08/04/11. 87 p. Advisor: Markus Endler.

Abstract: Mobile computing creates the need for applications to be adaptable according to the user's context. Specific user demands as well as changes in the computational context the mobile applications require clients to be able to adapt dynamically to suit the new scenario of execution. These adjustments should be appropriate and should maintain the quality of service avoiding failures or preventing degradation of application performance. This thesis proposes an extension of Kaluana middleware which provides a mechanism for selection of components for adaptative applications based on the notion of reconfigurations contracts. This selection is done also based on the notion of equivalence between their public interfaces and should consider the execution restrictions of the candidate components to be used in the adaptation in accord to the device's execution context to be used to evaluate the candidate components to the instantiation. This selection aims to maintain the compatibility of new components with the components already used as well as the execution context, i.e. the current status of the device's resources. Due to the notion of equivalence between the interface specifications of components, the application should be able to request a component through the interface of the requested service, avoiding the need to know the component's name or any other specific feature that creates a tie with the given implementation.


[11_MSc_rocha]
Allan Carlos Avelino ROCHA. Visualização volumétrica ilustrativa de malhas não estruturadas. [Title in English: Illustrative volume visualization for unstructured meshes]. M.Sc. Diss. Port. Presentation: 20/06/11. 54 p. Advisor: Waldemar Celles Filho.

Abstract: Scientific visualization techniques create images attempting to reveal complex structures and phenomena. Illustrative techniques have been incorporated to scientific visualization systems in order to improve the expressiveness of such images. The rendering of feature lines is an important technique for better depicting surface shapes and features. In this thesis, we propose to combine volume visualization of unstructured meshes with illustrative isosurfaces. This is accomplished by extending a GPU-based ray-casting algorithm to incorporate illustration with photic extremum lines, a type of feature lines able to capture sudden changes of luminance, conveying shapes in a perceptually correct way.


[11_MSc_reis]
André Luiz Castro de Almeida REIS. Editor de diagramas dirigido por metamodelos. [Title in English: Diagram editor driven by metamodels]. M.Sc. Diss. Port. Presentation: 26/08/11. 111 p. Advisor: Simone Diniz Junqueira Barbosa.

Abstract: The use of diagram editors has been very useful for creating design solutions in the area of human-computer interaction. They facilitate the use of modeling languages and provide control over the elements of the solution space, preventing the user from using an invalid lexical item of the chosen language. The elements are defined in a metamodel, which basically consists of a set of concepts with-in a given domain. As result, users gain speed and reliability in the process of creation. However, many editor do not guarantee that designed solution meets the language syntax. For this, we need an editor that, in addition to having control over the language symbols, also provides support for the use of models, going beyond graphical editing and also making use of the syntax rules defined in each metamodel. With this set of rules that define the from of language, the user may be warned of possible rule violations while building the solution. The rules discribe the syntax of the language through its grammar. To parse a diagram means to try to find a sequence of applications of rules that derive from a grammar or some representation of it. Considering this approach, this dissertation presents a study on diagram editors, and a metamodel-driven tool that allows the user to, by defining a metamodel, make use of a generic diagram editor for visuallanguages that can control the vocabulary and grammar of the created diagrams.Thus the goal of current research is to propose a tool that encompasses these solutions and is focused on visual languages commom in the area of human-computer interaction, such as MoLIC, CTT and statechart.


[11_MSc_carvalho]
Andréa Weberling CARVALHO. Criação automática de visões materializadas em SGBDs relacionais. [Title in English: Automatic creation of materialized views in relational DBMS]. M.Sc. Diss. Port. Presentation: 20/04/11. 83 p. Advisor: Edward Hermann Haeusler.

Abstract: As database applications become more complex, tuning database systems in order to improve query response times also become harder. One could consider materialized views, a relational database object that stores data resulting from specific queries, to obtain better performances. This dissertation proposes the automatic creation of materialized views. A non-intrusive architecture is used in order to keep DBMS source codes unchanged. There is a need to estimate the creation cost and the benefits obtained from the presence of the materialized views. Heuristics are proposed to help with the automatic decision upon creation of these materialized views for a given workload. Simulation tests for TPC-H benchmark and MS SQL Server DBMS are presented.


[11_MSc_monteiro]
Andrei Alhadeff MONTEIRO. Many-core fragmentation silmulation. [Title in Portuguese: Implementação de fragmentação em arquitetura de multi-processadores]. M.Sc. Diss. Port. Presentation: 17/08/11. 59 p. Advisor: Waldemar Celles Filho.

Abstract: A GPU-based computational framework is presented to deal with dynamic failure events simulated by means of cohesive zone elements. The work is divided into two parts. In the first part, we deal with pre-processing of the information and verify the effectiveness of dynamic insertion of cohesive elements in large meshes. To this effect, we employ a simplified topological data structured specialized for triangles. In the second part, we present an 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 extrinsic cohesive zone formulation resides on being able to dynamically adapt the mesh in a consistent way, inserting cohesive elements on fractured facets. In order to handle that, we extend the conventional data structure used in finite element code (based on element incidence) and store, for each element, references to the adjacent elements. To avoid concurrency on accessing shared entities, we employ the conventional strategy of graph coloring. In a pre-processing phase, each node of the dual graph (bulk element of the mesh) is assigned a color different to the colors assigned to adjacent nodes. In that way, elements of a same color can be processed in parallel without concurrency. All the procedures needed for the insertion of cohesive elements along fracture facets and for computing node properties are performed by threads assigned to triangles, invoking one kernel per color. Computations on existing cohesive elements are also performed based on adjacent bulk elements.


[11_MSc_souza]
Bruno de Figueiredo Melo e SOUZA. Modelos de fatoração matricial para recomendação de vídeos. [Title in English: Matrix factorization models for vídeo recommendation]. M.Sc. Diss. Port. Presentation: 29/08/11. 67 p. Advisor: Ruy Luiz Milidiú.

Abstract: Item recommendation from implicit feedback datasets consists of passively tracking different sorts of user behavior, such as purchase history, watching habits and browsing activities in order to improve customer experience through providing personalized recommendations that fits into users' taste. In this work we evaluate the performance of different matrix factorization models tailored for the recommendation task for the implicit feedback dataset extracted from Globo.com's video site's access logs. We propose treating the data as indication of a positive preference from a user regarding the video watched. Besides that we evaluated the impact of effects associated with either users or items, known as biases or intercepts, independent of any interactions and its time changing behavior throughout the life span of the data in the result of recommendations. We also suggest a scalable and incremental procedure, which scales linearly with the input data size. In trying to predict the intention of the users for consuming new videos our best factorization models achieves a RMSE of 0,0524 using user's and video's bias as well as its temporal dynamics.


[11_PhD_gadelha]
Bruno Freitas GADELHA. Uma abordagem de desenvolvimento de groupware baseada em linha de produto de software e modelo 3C de colaboracao. [Tittle in English: An approach for groupware development based on software product lines and the3C collaboration model].
Ph.D Thesis. Port. Presentation: 21/12/2011 101 p. Advisor: Hugo Fuks.

Abstract: In this thesis we explore software development on the context of groupware, specifically on supporting collaborative learning. Groupware development is not a trivial task given that technological and social issues are involved. Considering the technological issues, a huge amount of time is wasted on implementing infrastructure aspects leaving little time for implementation of innovative solutions on collaboration. Considering the social issues, we should take into account that group work is dynamic and that group composition changes over time. So, we developed a software product line for groupware based on the 3C Collaboration Model. The groupware derivation process starts with the formalization of the collaborative learning techniques in collaboration scripts. In order to support this collaboration process we developed the GroupwareBuilder, that reads the collaboration script and derives groupware tailored to the tasks described on the script. We made a functional evaluation and a case study. On the functional evaluation, we aimed on getting a proof of concept for GroupwareBuilder by deriving groupware for supporting the “Critical Debate” and “Buzz Groups” collaboration scripts. In order to analyze how GroupwareBuilder derives groupware from other collaborative learning techniques described by different teachers we made a case study. The main contribution of this thesis is an approach that enables the derivation of groupware and the customization of groupware in runtime from collaboration scripts written by the users, and not from a list of software requirements as used in other SPLs approaches.


[11_MSc_lima]
Bruno Seabra Nogueira Mendonça LIMA. Composer: aspectos não-funcionais em um ambiente de autoria para aplicações NCL. [Title in English: Composer: non-requirements aspects in an authoring environment to NCL applications]. M.Sc. Diss. Port. Presentation: 15/04/11. 81 p. Advisor: Luiz Fernando Gomes Soares.

Abstract: The chain of work present in the creation, development and transmission of hypermedia content includes several actors, environments and equipment, from the content creator, passing by the application developer all the way to the operator of the transmission service. Each of these actors is immersed in different work environments and has specific roles in the creation and editing of content being delivered to the final user.  Nowadays, even final users are demanding tools that can enrich their content. A single authoring tool cannot meet these requirements and different actors. Currently, there are authoring tools focused on satisfying a small subset of these actors. But even this small part is not fully satisfied, since these tools were built, most of the times, based only on functional requirements. This work discusses the importance of non-functional aspects in the development of new hypermedia authoring tools. This dissertation proposes an architecture that enables tools to meet the specific requirements of each actor in the process of creating hypermedia content. This architecture relies on extensibility, adaptability, performance and scalability. In order to test the proposal of this work we have developed an authoring tool for applications NCL (Nested Context Language) that is based on the proposed architecture. NCL was chosen because it is the standard language for the declarative system (Ginga-NCL) part of the Terrestrial Digital TV of Brazil ISDB-TB and ITU-T IPTV services. NCL allows the authoring of hypermedia documents in a simple and expressive form. 


[11_PhD_avila]
Bruno Tenório ÁVILA. Compressão de números naturais, sequência de bits e grafos. [Title in English: Compression of natural numbers, sequence of bits and graphs]. Ph.D. Thesis. Port. Presentation: 16/09/11.  100 p. Advisor: Eduardo Sany Laber.

Abstract: This thesis addresses the problems of compression for the following data types: numbers, sequence of bits and webgraphs. For the problem of compression of a sequence of bits, we demonstrate the relationship between merge algorithms and binary source coders. Then, we show that the algorithms binary merge, recursive merge and probabilistic merge, generate respectively an entropy coder based runlengths encoded with the Rice code, the interpolative binary coder and the random Rice coder, which is a new variant of the Rice code. For the problem of webgraph compression, we propose a new compact representation for webgraphs, entitled w-tree, built specifically for external memory (disk), being the first one in this genre. We also propose a new type of layout designed specifically for webgraphs, entitled scaled layout.In addition, we show how to build a cache-oblivious layout to explore the hierarchy of memories, being the first of its kind. We offer several types of queries that can be performed and it is the first representation to support batched random read query execution and advanced query optimization, including in main memory. Finally, we performed a series of experiments showing that the w-tree provides compression rates and running times competitive with other compact representations for main memory. Therefore, we demonstrate empirically the feasibility of a compact representation for external memory in practice, contrary to the assertion of several researchers.


[11_MSc_marques]
Carlos Roberto MARQUES JUNIOR. Configuração colaborativa de linha de produtos de software. [Title in English: Collaborative configuration of software product line]. M.Sc. Diss. Port. Presentation: 07/04/10. 98 p. Advisor: Carlos José Pereira de Lucena.

Abstract: Product configuration is a key activity for enabling mass customization. It corresponds to tailoring a software application from a software product line, respecting individual customer requirements. In practice, the product configuration activity is challenging, mainly because it involves numerous stakeholders with different expertise making decisions. Several works claim to decompose the configuration activity into pre-arranged stages, whereby stakeholders can make their decisions in a separated way and coherent fashion. However, due to the decentralization in geography of the knowledge required in each stage and that customer requirements could be imprecise and change frequently, the proposed solutions do not apply. To address these issues, this work proposes a dynamic and collaborative product configuration approach based on the personal assistant metaphor. Personal assistants in a cooperative way coordinate stakeholders´ decisions and proactively perform tasks, such as reasoning about the ramifications of decisions, integrating distributed decisions and resolving divergent requirements. A tool prototype, called Libertas, and two case studies that evaluate the applicability of our approach are also presented. The first case study analyzes the configuration process to configure an operational system to support the business requirements of an enterprise. The second one addresses a scenario of a software product line for web portals.
 

[11_MSc_campos]
Daniel de Vasconcelos CAMPOS. SisApC2: uma estratégia baseada em sistemas computacionais móveis para apoiar atividades de comando e controle. [Title in English: SisApC2: A strategy based on mobile computer systems to support Command and Control activities]. M.Sc. Diss. Eng. Presentation: 29/03/11. 88 p. Advisors: Marcelo Gattass and Roberto de Beauclair Seixas.

Abstract: The Command and Control theory of John Boyd, a 20th century military strategist, is a powerful tool not only for the military, but also for any civilian activities that require monitoring people, vehicles, boats, or any other elements of interest. Nowadays computer graphics techniques are important for the OODA loop (Observe–Orient–Decide–Act), especially on observing and orienteering steps. They make these steps more efficient and less prone to errors. The implementation of Command and Control techniques with current technology is not a simple task. It involves the implemantation of graphical, mobile and distributed components. Each of these components can be implemented with a variety of technologies, that are often incompatible with each other. This dissertation proposes a low cost framework capable of effectively support the implemantation of a Command and Control computer system. The framework is based on open source reliable technologies and has proven versatile for different types of applications. The dissertation also presents the implementation of two different applications with the proposed technology to support its evaluation.


[11_MSc_conceicao]
Diêgo Bispo Conceição. Simulação e estratégias de negociação de ações com agentes de software. [Title in English: Simulation and Stock Trading Strategies with Software Agents]. M.Sc. Diss. Port. Presentation: 02/09/11. 82 p. Advisor: Carlos José Pereira de Lucena.

Abstract: The financial market has presented significant growth in the automation of decisions and execution of strategies that can achieve good returns from investments. Consequently, the need for an increasingly robust and reliable environment, allowing to analyze different investment strategies, has increased. Based on this need, this work presents "A Multi-Agent System Framework For Automated Stock Exchange Simulation" (FrAMEX), which allows the creation of different simulators for the financial market based on the paradigm of software agents. Intraday and interday simulators created from FrAMEx are presented in the document. Besides the analysis of different investment strategies used in such environments and executed by agents run from investors. Since these agents achieved good performances in their executions, they participated in two versions of the MASSES competition. Thus, the description of the performance of each agent developed is also presented.

 
[11_MSc_cardoso]
Eduardo Teixeira
CARDOSO. Efficient methods for information extraction in news webpages. [Title in Portuguese: Métodos eficientes para extração de informação em páginas de notícias]. M.Sc. Diss. Port. Presentation: 24/08/11. 58 p. Advisor: Eduardo Sany Laber.

Abstract: We tackle the task of news webpage segmentation, specifically identifying the news title, publication date and story body. While there very good results in the literature, most of them rely on webpage rendering, which is a very time-consuming step. We focus on scenarios with a high volume of documents, where a short execution time is a must. The chosen approach extends our previous work in the area, combining structural properties with hints of visual presentation styles, computed with a faster method than regular rendering, and machine learning algorithms. In our experiments, we took special attention to some aspects that are oten overlooked in the literature, such as processing time and the generalization of the extraction results for unseen domains. Our approach has shown to be about an order of magnitude faster than an equivalent full rendering alternative while retaining a good quality of extraction.

 
[11_MSc_miranda]
Fabio Markus Nunes MIRANDA. Volume rendering of unstructured hexahedral meshes. [Title in Portuguese: Renderização volumétrica de malha não estruturada de hexaedros]. M.Sc. Diss. Port. Presentation: 05/09/11. 47 p. Advisor: Waldemar Celles Filho.

Abstract: Important engineering applications use unstructured hexahedral meshes for numerical simulations. Hexahedral cells, when compared to tetrahedral ones, tend to be more numerically stable and to require less mesh refinement. However, volume visualization of unstructured hexahedral meshes is challenging due to the trilinear variation of scalar fields inside the cells. The conventional solution consists in subdividing each hexahedral cell into five or six tetrahedra, approximating a trilinear variation by an inadequate piecewise linear function. This results in inaccurate images and increases the memory consumption. In this thesis, we present an accurate ray-casting volume rendering algorithm for unstructured hexahedral meshes. In order to capture the trilinear variation along the ray, we propose the use of quadrature integration. We also propose a fast approach that better approximates the trilinear variation to a series of linear ones, considering the points of minimum and maximum of the scalar function along the ray. A set of computational experiments demonstrates that our proposal produces accurate results, with reduced memory footprint. The entire algorithm is implemented on graphics cards, ensuring competitive performance.

 
[11_MSc_pina]
Felipe Freixo PINA. Aplicações de DHT em sistemas de computação distribuída. [Title in English: Utilization of DHT in distributed computing systems]. M.Sc. Diss. Port. Presentation: 29/08/11. 63 p. Advisor: Noemi de La Rocque Rodriguez.

Abstract: P2P architectures are recognized for decentralization and incentive for the cooperation among nodes. These characteristics allow for fault tolerance and resource distribution among the nodes (by replication) to systems based on the P2P architecture. Systems based in P2P networks built using the DHT technique are scalable. Since this architecture is commonly used in content distribution systems, in this work we investigate the utilization of the DHT technique in distributed computing systems, where the shared resources are the node’s computational power. Four routing protocols were analyzed to identify the most appropriated for use in distributed computing systems and applied the group concept to archive fault tolerance and resource distribution among nodes.

 
[11_PhD_viana]
Francisco Henrique de Freitas VIANA. Modelos e algoritmos para o team orienteering problem. [Title in English: Models and algorithms to the team orienteering problem]. Ph.D. Thesis. Port. Presentation: 12/09/11. 99 p. Advisors: Marcus Vinicius Soledade Poggi de Aragão and Eduardo Uchoa Barboza.

Abstract: Team Orienteering Problem is a vehicle routing problem on a graph with durations associated to the arcs and profits assigned to visiting the vertices. In this problem, a fleet with a fixed number of identical vehicles performs the visitations and there is a limited total duration for the routes to be ended up. Each vertex can be visited at most once and the solution does not have the obligation to visit all vertices, due to the constraint that limits the maximum duration of routes. The goal of the problem is to maximize the total profit gathered by all routes. In this work, two approaches have been proposed: an exact and a heuristic one. In the exact approach, we have developed an arc based formulation and an extended formulation where each arc has an extra index. This index represents the departure time of a vehicle using an arc. Through transformations on the extended formulation, we have obtained a formulation, whose relaxation - the restricted master problem - is solved using the column generation technique. A dynamic programming algorithm solves the column generation subproblem in pseudo-polynomial time. This algorithm generates non-elementary routes that allow subcycles. In order to cut off the subcycles, a new class of inequalities called min cut has been proposed.We have applied a Branch-Cut-and-Price (BCP) algorithm. This allowed finding some new upper bounds. The exact approach has achieved competitive results compared to the best exact algorithm has already proposed to this problem. In the heuristic approach, besides a kopt neighborhood, we have also exploited an ellipsoidal search that adds a new cut constraint to the formulation of Branch-Cut-and-Price algorithm. This new cut reduces the space search to a neighborhood around a known set of solutions. This search is used as a crossover operator that runs all iterations of a evolutive algorithm. This approach converges in a reasonable computational time and finds optimal or near optimal solutions for some instances in the literature.

 
[11_PhD_abraham]
Frederico Rodrigues ABRAHAM. Visualização de modelos massivos de reservatórios naturais de petróleo. [Title in English: Visualization of complex natural black oil reservoir models]. Ph.D. Thesis. Port. Presentation: 12/09/11. 122 p. Advisor: Waldemar Celles Filho.

Abstract: Recent advances in parallel architectures for the numerical simulation of natural black oil reservoirs have allowed the simulation of very discretized domains. As a consequence, these simulations produce an unprecedented volume of data, which must be visualized in 3D environments for carefull analysis and inspection of the model. Conventional scientific visualization techniques of such very large models are not viable, creating a demand for the development of scalable visualization solutions. The need for the visualization of such complex data introduces several computational issues which must be addressed in order to achieve interactive rendering rates, such as the impossibility of storing the entire data in main memory. There are two main research areas which propose solutions for the visualization of models with such magnitude: distributed rendering and multi-resolution techniques. This work proposes solutions for the visualization of massively complex reservoir models in each of these research areas, and a discussion over the advantages and limitations of each solution is made. In the first part of the work, we propose a distributed system based on a sort-last approach for the rendering of such models in PC clusters, where each PC is equipped with multiple GPUs. Given an efficient use of the available GPUs, combined with a pipelined implementation and the use of partial image compositions on the cluster nodes, our proposal tackles the scalability issues that arise when using mid-to-large GPU clusters. The second part of the work brings the proposal of a hierarchical multi-resolution structure of black oil reservoir meshes, with a new simplification algorithm designed specifically for such meshes. The hierarchical structure brings some new approaches in relation to related work, doing a much less conservative projected error estimation. We propose a minimum refresh rate guarantee strategy for our multiresolution rendering, which is the main goal for such systems. Afterwards, we introduce a proposal for the rendering of data associated with the original reservoir mesh mapped over the simplified meshes, such as the original model grid wireframe and reservoir properties. This proposal guarantees an independence between the multi-resolution structure and the properties generated by a simulation, which guarantees the reuse of the structure among several simulations of the same model. Experimental results demonstrate the effectiveness of the proposed solutions.


[11_PhD_faustino]
Geisa Martins FAUSTINO. Um método baseado em mineração de grafos para segmentação e contagem de clusters de máximos locais em imagens digitais. [Title in English: A graph-mining based method for segmentation and counting of local maximum clusters in digital images. Ph.D. Thesis. Port. Presentation: 08/04/11. 79 p. Advisors: Marcelo Gattass and Carlos José Pereira de Lucena.

Abstract: A grayscale image can be viewed as a topological surface and this way, objects of interests may appear as peaks (sharp mountains), domes (smooth hills) or valleys (V- or U-shaped). Generally, the dome top presents more than one local maximum. Thus, it can be characterized by a local maximum cluster. Segmenting objects individually in images where they appear partially or totally fused is a problem which frequently may not be solved by a watershed segmentation or a basic morphological processing of images. Other issue is counting similar objects in images segmented beforehand. Counting them manually is a tedious and time-consuming task, and its subjective nature can lead to a wide variation in the results. This work presents a new method for segmenting and counting of local maximum clusters in digital images through a graph-based approach. Using the luminance information, the image is represented by a region adjacency graph and a graph-mining algorithm is applied to segment the clusters. Finally, according to image characteristics, a graph-clustering algorithm can be added to the process to improve the final result. The object counting step is a direct result from the mining algorithm and the clustering algorithm, when the latter is applied. The proposed method is tolerant to variations in object size and shape and can easily be parameterized to handle different image groups resulting from distinct objects. Tests made on a database with 262 images, composed of photographs of objects (group 1) and embryonic stem cells under fluorescence microscopy images (group 2), attest the effectiveness and quality of the proposed method as for segmentation and counting purpose. The images form group 1 processed by our method were checked by the author and those ones from group 2 by the specialists from the Institute of Biomedical Sciences at UFRJ. For these images we obtained an average F-measure of 85.33% and 90.88%, respectively. Finally, a comparative study with the widely used watershed algorithm was done. The watershed achieved an average F-measure of 74.02% e 78.28% for groups 1 and 2, respectively, against 85.33% e 91.60% obtained by our method.


[11_MSc_lima]
Guilherme Augusto Ferreira LIMA. Eliminando redundâncias no perfil NCL EDTV. [Title in English: Eliminating redundancies from NCL EDTV Profile]. M.Sc. Diss. Port. Presentation: 10/06/11.  70 p. Advisor: Luiz Fernando Gomes Soares.

Abstract: The implementation of a NCL presentation engine, or formatter, is a complex task. This complexity is mainly due to the semantic distance between NCL documents, high-level declarative specifications, and the API used by the formatter to present them, in most cases low-level and imperative. The greater the distance, the greater is the complexity of this mapping and, consequently, of its implementation, which is more likely to become ineficient and bug-prone. This work presents a new NCL profile, called NCL Raw, which eliminates most of the redundancies present in EDTV | the main profile of NCL 3.0 | and, in a certain way, reduces the distance between the documents and the machine. Raw profile captures only EDTV s essential concepts, which in turn can be used to simulate the whole language defined by EDTV itself. In other words, we can use the Raw profile as a simpler intermediate language to which EDTV documents can be converted before being presented. This dissertation discusses alternative architectures for NCL converters and presents the implementation of a document converter (from EDTV to Raw).


[11_MSc_ferreira]
Guilherme Carlos De Napoli Ferreira. A machine learning approach for Portuguese text chunking. [Title in Portuguese: Uma abordagem de aprendizado de máquina para segmentação textual no português]. M.Sc. Diss. Port. Presentation: 13/04/11. 63 p. Advisor: Ruy Luiz Milidiú.

Abstract: Text chunking is a very relevant Natural Language Processing task, and consists in dividing a sentence into disjoint sequences of syntactically correlated words. One of the factors that highly contribute to its importance is that its results are used as a significant input to more complex linguistic problems. Among those problems we have full parsing, clause identification, dependency parsing, semantic role labeling and machine translation. In particular, Machine Learning approaches to these tasks greatly benefit from the use of a chunk feature. A respectable number of effective chunk extraction strategies for the English language has been presented during the last few years. However, as far as we know, no comprehensive study has been done on text chunking for Portuguese, showing its benefits. The scope of this work is the Portuguese language, and its objective is twofold. First, we analyze the impact of different chunk definitions, using a heuristic to generate chunks that relies on previous full parsing annotation. Then, we propose Machine Learning models for chunk extraction based on the Entropy Guided Transformation Learning technique. We employ the Bosque corpus, from the Floresta Sintá(c)tica project, for our experiments. Using golden values determined by our heuristic, a chunk feature improves the F!=1 score of a clause identification system for Portuguese by 6.85 and the accuracy of a dependency parsing system by 1.54. Moreover, our best chunk extractor achieves a F!=1 of 87.95 when automatic part-of-speech tags are applied. The empirical findings indicate that, indeed, chunk information derived by our heuristic is relevant to more elaborate tasks targeted on Portuguese. Furthermore, the effectiveness of our extractors is comparable to the state-of-the-art similars for English, taking into account that our proposed models are reasonably simple.


[11_MSc_nunes]
Gustavo Bastos NUNES. Explorando aplicações que usam a geração de vértices em GPU. [Title in English: Exploring applications that use vertex generation on GPU]. M.Sc. Diss. Port. Presentation: 16/08/11.  111 p. Advisors: Alberto Barbosa Raposo and Bruno Feijó.

Abstract: One of the main bottlenecks in the graphics pipeline nowadays is the memory bandwidth available between the CPU and the GPU. To avoid this bottleneck, programmable features were inserted into the video cards. With the Geometry Shader launch it is possible to create vertices in the GPU, however, this pipeline stage has a low performance. With the new graphic APIs (DirectX11 and OpenGL4) a Tessellator stage that allows massive vertex generation inside the GPU was created. This dissertation studies this new pipeline stage, as well as presents classic algorithms (PN-Triangles and Phong Tessellation) that were originally designed for CPU and proposes new algorithms (Tubes and Terrain rendering in the GPU) that takes advantage of this new paradigm.


[11_MSc_roenick]
Hugo Roenick. Um modelo de componentes de software com suporte a múltiplas versões. [Title in English: A software component model with support for multiple versions]. M.Sc. Diss. Port. Presentation: 04/03/11. 66 p. Advisor: Renato Fontoura de Gusmão Cerqueira.

Abstract: Several software component models for distributed systems have been proposed by the industry and academy, such as Fractal, CCM, COM, OpenCOM, LuaCCM, and SCS. One of the greatest advantages of component-based development is the better support for independent extensibility. However, managing multiple versions of components is still a challenge, especially when it’s not possible to update all the system’s components at the same time. Because of that, different versions of the same component interface shall be required to coexist in the same system. In this work, we try to identify the key points to support multiple versions of component interfaces and propose a model that offers this support. To evaluate the proposed model, we will extend the SCS component system to support it. Based on the evolution historic of different SCS components used in a real application, we will conduct experiments with the new version of SCS to verify the effectiveness of the proposed model.


[11_MSc_monteiro]
Ingrid Teixeira MONTEIRO. Acessibilidade por diálogos de mediação: desenvolvimento e avaliação de um assistente de navegação para a Web. [Title in English: Accessibility by mediation dialogues: development and evaluation of a Web navigation helper]. M.Sc. Diss. Port. Presentation: 16/03/10.  198 p. Advisor: Clarisse Sieckenius de Souza.

Abstract: Web accessibility is one of the big challenges in the Computer Science research. There are many initiatives that aim to improve systems, in order that users with disabilities and other special needs have plain access to information and service available on the Internet. One of them is the system presented here, WNH, a Web Navegation Helper, created to support users with special needs do activities on the Internet, through mediation dialogs, previously development, with a specialized editor, by users interested in helping these people. We present in the text, the description of the developed tools (the editor and the helper) and the analysis of three exploratory studies done, before and after the system development. We also show how the experiments revealed social and cultural aspects of brazilian society, that are relevant to the WNH design, and we show how they changed our initial vision of the system. There was necessary to rethink the tool, in order to take account on the cultural variable in its development and evaluation.


[11_MSc_vieira]
Jerônimo Sirotheau de Almeida EICHLER. Uma arquitetura para inferência de atividades de usuário de computação móvel. [Title in English: An architecture for inference of activities of mobile computing users]. M.Sc. Diss. Port. Presentation: 12/04/11 67 p. Advisor: Karin Koogan Breitman.

Abstract: The ubiquitous computing combined with the advance of sensor technology creates a scenario in which the integration of several computing resources is used to keep a set of services and features available to the user whenever necessary. A particular trend in this area is the activity based systems, i.e., systems that are aware of the activity played by the user. In these systems, inference engine is essential to recognize user’s actions and allow the systems to adapt its behavior according to user’s actions. Though, the development of this type of systems is not a trivial task as the high rate of information exchanged brings challenges related to privacy, performance and information management. In this work we propose an architecture for activity inference systems. To achieve this goal, we define a set of components that perform important roles in the inference process. Finally, to show the feasibility of this approach, we designed, implemented and evaluated a system based on the proposed architecture. 


[11_MSc_vieira]
José Geraldo de Sousa JUNIOR. Uma arquitetura para aplicações dinâmicas NCL baseadas em famílias de documentos. [Title in English: An architecture for dynamic NCL applications based on document families
]. M.Sc. Diss. Port. Presentation: 09/11/11. 72 p. Advisor: Luiz Fernando Gomes Soares.

Abstract: The presentation of dynamic hypermedia applications may be seen as a recursive authoring process, in which applications are recreated during presentation time, whenever content changes are triggered by interactions between the presentation engine and other entities such as users, imperative objects, external applications, etc. In some scenarios of dynamic hypermedia applications, it is possible to identify a hypermedia composition pattern that remains consistent even after the document is recreated. This kind of applications is common, for instance, in an Interactive Digital Television environment. The presence of such a pattern throughout the presentation of an application for Interactive Digital Television allows the establishment of an
analogy between recreating documents dynamically and authoring applications through a template-driven authoring method. Using the latter, the authoring process is conducted by “filling” gaps left by a template that represents the hypermedia composition pattern of an application. Analogously, in the dynamic document re-creation, the module that processes document updates fulfills the role of “filling” the templates gaps. The main goal of the present work is to define an architecture, inspired by this analogy, to structure NCL applications that can be dynamically refactored and that remain conform to their respective templates. Nested Context Language (NCL) is the language of Brazilian System of Digital Terrestrial Television applications. In order to validate the proposal, an application that captures a real scenario and an authoring tool for specify graphically document filling was developed.


[11_MSc_vieira]
Lourival Pereira VIEIRA NETO. Lunatik: scripting de kernel de sistema operacional com Lua. [Title in English: Operating system kernel scripting with Lua] M.Sc. Diss. Port. Presentation: 12/04/11  69 p. Advisor: Roberto Ierusalimschy.

Abstract: There is a design approach to improve operating system flexibility, called extensible operating system, that supports that operating systems must allow extensions in order to meet new requirements. There is also a design approach in application development that supports that complex systems should allow users to write scripts in order to let them make their own configuration decisions at run-time. Following these two design approaches, we have built an infrastructure that allows users to dynamically load and run Lua scripts into operating system kernels, improving their flexibility. In this thesis we present Lunatik, our scripting subsystem based on lua, and show a real usage scenario in dynamically scaling CPU frequency and voltage. Lunatik is currently implemented both for NetBSD and Linux.


[11_PhD_salgado]
Luciana Cardoso de Castro SALGADO. Cultural viewpoint metaphors to explore and communicate cultural perspectives in cross-cultural HCI design. [Title in Portuguese: Metáforas de perspectivas culturais para exploração e comunicação da diversidade cultural no design de IHC] Ph.D. Thesis. Eng. Presentation: 04/04/11 228 p. Advisor: Clarisse Sieckenius de Souza.

Abstract: More than ever before, today one of the challenges for interaction design is the development of systems aiming to attend to the needs and expectations of people with different cultural and social backgrounds. The most widely used perspective in cross-cultural design is internationalization-localization. Internationalization is the process of separating the core functionality code from system's interface specifics (e.g. text language, measures, etc.). With localization, the interface is customized for a particular audience (through) language translation, cultural markers and even technical features, for instance). The result of internationalization and localization is to conceal or neutralize cultural differences among different user communities and contexts of use. We are, however, interested in another situatuin: one where the design intent is virtually the opposite: to expose and explore cultural diversity. This is the case, for instance, when the purpose of the designed system is to simulate users to make contact with a foreign culture. This thesis provides new knowledge to help HCI designers communicate their intent when they want to promote the users' contact with cultural diversity. We present five cultural viewpoint metaphors (CVM) to support reasoning and decision-making about intercultural experience dimensions. The metaphors derive from empirical studies applying Semiotic Engineering to analyze and re-design cross-cultural systems interfaces. In order to investigate if and how CVM actually support HCI professional/practitioners at design and evaluation time, we carried out an extensive case study to assess how CVM can be used in design and evaluation activities. We found that CVM played an important role in early design stages, helping designers to reason effectively about intercultural experiences while determining which cultural perspective they want to adopt. Furthermore, CVM features provided a rich epistemic grid where to consistency of design choices stands out more clearly.


[11_MSc_gomes]
Luciana da Silva Almendra GOMES. Proveniência para workflows de Bioinformática. [Title in English: Provenance for Bioinformatics workflows]. M.Sc. Diss. Port. Presentation: 27/04/11. 104 p. Advisor: Edward Hermann Haeusler.

Abstract: Many scientific experiments are designed as computational workflows, which can be implemented using traditional programming languages. In the Bioinformatics domain ad-hoc scripts are often used to build workflows. Scientific Workflow Management Systems (SWMS) have emerged as an alternative to those scripts. One particular SWMS feature that has received much attention by the scientific community is the automatic capture of provenance data. These allow users to track which resources and parameters were used to obtain the results, among many other required information to validate and publish an experiment. In the present work we have elicited some data provenance challenges in the SWMS context, such as (i) the heterogeneity of data representation schemes that hinders the understanding and interoperability; (ii) the storage of consumed and produced data and (iii) the reproducibility of a specific execution. These challenges have motivated the proposal of a data provenance conceptual scheme for workflow representation. We have implemented an extension of a particular SWMS system (Bioside) to include provenance data and store them using the proposed conceptual scheme. We have focused on some requirements commonly found in bioinformatics workflows.


[11_PhD_valente]
Luis Paulo Santos VALENTE. A methodology for conceptual design of pervasive mobile games. [Title in Portuguese: Um modelo de domínio para jogos pervasivos móveis]. Ph.D. Thesis. Eng. Presentation: 09/09/11. 242 p. Advisor: Bruno Feijó.

Abstract: Pervasiveness can be recognized in game playing every time the boundaries of playing expand from the virtual (or fictional) world to the real world. Sensor technologies, mobile devices, networking capabilities, and the internet make pervasive games possible. In the present work, we consider “pervasive mobile games” as context-aware games that necessarily use mobile devices. Also we consider that smartphones are the main driver to fulfill the promises of pervasive game playing. As far as we are aware, this is the first general work on pervasive mobile game design. This thesis proposes a methodology to support the conceptual design stage of pervasive mobile games. The main contributions of this research work are twofold: [1] A novel list of prominent features of pervasive games, identified from game projects and the literature, and checklists for each feature. This feature list (and corresponding checklists) can be used to spark novel game ideas, and to help in discovering functional and non-functional requirements for pervasive mobile games. [2] A domain specific language to help in specifying activities in pervasive mobile games that use mobile devices, sensors, and actuators as the main interface elements.With the proposed methodology, designers can discuss, identify, verify, and apply important features of pervasive mobile games. Also, due to the “lightweight” nature of the methodology, designers can easily catch the “big picture” of the games by keeping focused on the intents of the game activities, and not getting lost in the source code.


[11_MSc_belchior]
Mairon de Araújo BELCHIOR. Modelo de controle de acesso no projeto de aplicações na Web semântica. [Title in English: An access control model for the design of semantic Web applications].
M.Sc. Diss. Port. Presentation: 10/10/2011. 110 p. Advisor: Daniel Schwabe

Abstract: The Role-based Access Control (RBAC) model provides a way to manage access to information of an organization, while reducing the complexity and cost of security administration in large networked applications. Currently, several design method of Semantic Web (and Web in general) applications was proposed, but none of these methods produces an specialize and integrated model for describing access control policies. The goal of this dissertation is to integrate the access control in design method of Semantic Web applications. More specifically, this work presents an extension of SHDM method (Semantic Hypermedia Design Method) in order to include RBAC model and an rule based policy Model integrated with the other models of this method. SHDM is a model-driven approach to design web applications for the semantic web. A modular software architecture was proposed and implemented in Synth, which is an application development environment according to SHDM method.


[11_MSc_azambuja]
Marcello de Lima AZAMBUJA. A cloud computing architecture for large scale video data processing. [Title in Portuguese: Uma arquitetura em nuvem para processamento de video em larga escala]. M.Sc Diss. Eng. Presentation: 31/08/11. 61 p. Advisor: Karin Koogan Breitman

Abstract: The advent of the Internet poses great challenges to the design of public submission systems as it eliminates traditional barriers, such as geographical location and cost. With open global access, it is very hard to estimate storage space and processing power required by this class of applications. In this thesis we
explore cloud computing technology as an alternative solution. The main contribution of this work is a general architecture in which to built open access, data intensive, public submission systems. A real world scenario is analyzed using this architecture for video processing.

[11_PhD_metello]
Marcelo Gomes METELLO. Process-oriented modeling and simulation for serious games. [Title in Portuguese: Modelagem e simulação orientadas a processos para jogos sérios]. Ph.D Thesis. Eng. Presentation: 31/08/11. 163 p. Advisor: Marco Antonio Casanova.

Abstract: This thesis focuses on serious games that simulate realistic situations. The objectives of such games go beyond mere entertainment to fields such as training, for example. Since other areas of Computer Science provide methods and tools for simulating and reasoning about real situations, it is highly desirable to use them in this kind of serious games. This thesis introduces a new framework on which simulation techniques from different areas, such as modeling and simulation, geographic information systems and multi-agent systems, can be integrated into a serious game architecture. The proposed solution resulted in the conception of a novel simulation modeling paradigm, named process-oriented simulation (POS), which combines different aspects of the more traditional object-oriented simulation (OOS) and agent-oriented simulation (AOS) paradigms. The main idea of POS is the separation between state and behavior of the entities involved in the simulation. This characteristic favours the modularization of complex behaviors and the integration of different and interfering simulation models into a single simulation. Based on the POS paradigm, a discrete-event simulation formalism named Process-DEVS was developed as an extension of the well-known DEVS simulation formalism. Some formalisms, such as workflows and cell space processes, were mapped to Process-DEVS and tested in the implementation of two systems: an emergency training game and a contingency planning system, both designed for the oil and gas industry.


[11_MSc_pessoa]
Marcos Borges PESSOA. Geração e execução automática de scripts de teste para aplicações web a partir de casos de uso direcionados por comportamento. [Title in English: Automatic generation and execution of test scripts for web applications from use case driven by behavior]. M.Sc Diss. Port. Presentation: 30/08/11. 99 p. Advisor: Arndt von Staa.

Abstract: This work aims at exploring the software requirements, described in the form of use cases, as an instrument to support the automatic generation and execution of functional tests, in order to automatically check if the results obtained in the tests generated and executed are in accordance with specified. The establishes a process and a tool for documenting use cases and automatically generating and executing test scripts that verify the behavior of web applications. The content of the use case, especially the flow of events (main and alternative), is structured in accordance to a "behavior model" that stores the test data and generates input for a browser testing tool. In this work, we have used the Selenium tool to automate the interaction with the browser. The assessment of our approach involved the application of the process and generating tool in real systems, comparing the results with other techniques applied in the same systems.


[11_MSc_salcedo]
Matheus SALCEDO. Gestão do conhecimento no gerenciamento de serviços de TI: uso e avaliação. [Title in English: Knowledge management for IT service management: usage and evaluation]. Ms.C. Diss. Port. Presentation: 11/01/11. 123 p. Advisor: Daniel Schwabe.

Abstract: In today's economy, the competitive advantage of companies can be directly linked to its ability to utilize the knowledge held by its members. However, to add value to an organization, this knowledge must be shared. Thus, the organization's ability to integrate and apply the expertise of its workforce is essential to achieve and maintain an organizational competitive advantage. The knowledge management systems help to create, collect, organize and disseminate knowledge. However, these systems have limitations, such as the difficulty of integrating knowledge from different sources, usually because of the lack of semantics in his descriptions. The main objective of this dissertation is to study the technological limitations of existing knowledge management systems and propose solutions through the adoption of Semantic Web formalisms. To achieve this goal, is used as a study case the knowledge management system in production of a large Brazilian company, which supports the operation of its IT infrastructure. This study demonstrates that approach can add additional semantics to existing data, integrating previously isolated databases, creating a better operating result.


[11_MSc_bomfim]
Mauricio Henrique de Souza BOMFIM. Um método e um ambiente para o desenvolvimento de aplicações na Web Semântica. [Title in English: A method and environment for Semantic Web application development]. M.Sc. Diss. Port. Presentation: 27/01/11. 196 p. Advisor: Daniel Schwabe.

Abstract: The growing availability of data and ontologies according to the Semantic Web standards has led to the need of methods and tools for applications development that take account the use and availability of the data distributed according to these standars. The goal of this dissertation is to present a method, including processes and models, and an environment for the Semantic Web applications development. More specifically, this work shows the evolution of the SHDM (Semantic Hypermedia Design Method), which is a method for the Semantic Web hypermedia application development and the Synth, which is an environment to build applications designed according to the SHDM.


[11_PhD_serrano]
Maurício SERRANO. Desenvolvimento intencional de software transparente baseado em argumentação. [Title in English: Intentional development of transparent software based on argumentation]. Ph.D. Thesis. Port. Presentation: 23/08/11. 283 p. Advisor: Julio Cesar Sampaio do Prado Leite.

Abstract: Transparency is a critical quality criterion to modern democratic societies. As software permeates society, transparency has become a concern to public domain software, as eGovernment, eCommerce or social software. Therefore, software transparency is becoming a quality criterion that demands more attention from software developers. In particular, transparency requirements of a software system are related to non-functional requirements, e.g. availability, usability, informativeness, understandability and auditability. However, transparency requirements are particularly difficult to validate due to the subjective nature of the involved concepts. This thesis proposes a transparency-requirements-driven intentional development of transparent software. Transparency requirements are elicited with the support of a requirements patterns catalog, relatively validated by the stakeholders through argumentation and represented on intentional models. Intentional models are fundamental to software transparency, as they associate goals and quality criteria expected by the stakeholders with the software requirements. The goals and quality criteria also justify the decisions made during software development. A system was implemented as an intentional multi-agents system, i.e., a system with collaborative agents that implement the Belief-Desire-Intention model and that are capable of reasoning about goals and quality criteria. This thesis discusses important questions to the success of our approach to the development of transparent software, such as: (i) forward and backward traceability; (ii) a fuzzy-logic based reasoning engine for intentional agents; (iii) the application of an argumentation framework to relatively validate transparency requirements through stakeholders’ multi-party agreement; and (iv) collaborative pre-traceability for intentional models based on social interactions. Our ideas were validated through case studies from different domains, such as ubiquitous computing and Web applications.


[11_PhD_serrano]
Milene SERRANO. Reuse-oriented approach for incremental and systematic development of intentional ubiquitous applications. [Title in English: Abordagem orientada à reutilização de software para desenvolvimento incremental e sistemático de aplicações ubíquas intencionais]. Ph.D. Thesis. Eng. Presentation: 31/03/11. 228 p. Advisor: Carlos José Pereira de Lucena.

Abstract: Ubiquitous applications are embedded in itelligent environments integrated into the physical world and composed of users with different preferences, heterogeneous devices and several content and service providers. Moreover, they focus on offering services and contents anywhere and at any time by assisting the users in their daily activities without disturbing them. Based on this ideliazed world, the "anywhere and at any time" paradigm poses some challenges for the Software Engineering community, such as: device heterogeneity, distributed environments, mobility, user satisfaction, content adaptability, context awareness, privacy, personalization, transparency, invisibility and constant evolution of technological trends. In order to deal with these new technological challenges, we propose a Reuse-Oriented Approach for Incremental and Systematic Development of Intentional Ubiquitous Applications. We have chosen two main goals that drive our research in this thesis: (i) the construction of reuse-oriented support sets based on an extensive investigation of ubiquitous applications and the Intentional-Multi-Agent Systems paradigm - i.e. Development for Reuse; and (ii) the incremental and systematic development of International-Multi-Agent-Systems-driven ubiquitous applications based on the reuse-oriented approach - i.e. Development with Reuse. Some contributions of our work are: (i) a reuse-oriented architecture centered on support sets - i.e. building blocks mainly composed of conceptual models, frameworks, patterns and libraries - obtained from the Domain Engineering of Ubiquitous Applications; (ii) a reuse-oriented Ubiquitous Application Engineering for incremental and systematic development of intentional ubiquitous applications centered on the proposed building blocks; (iii) a reasoning engine focused on fuzzy conditional rules and the Belief-Desire-Intention model to improve the agents' cognitive capacity; (iv) a specific mechanism based on intentional agents to deal with privacy issues by balancing privacy and personalization as well as transparency and invisibility; (v) a catalogue that graphically presents the main ubiquitous non-functional requirements, their interdependencies and ways to opreationalize them based on the combination of thaditional and emergent technologies; (vi) ontologies to allow the dynamic construction of interfaces and to improve the communication and inter-operability of software agents; and (vii) a dynamic database model to store and retrive the ubiquitous profiles (e.g. user, device, network and contract profiles) by improving the data management "on the fly". The proposed approach was evaluated by developing different ubiquitous applications (e.g. e-commerce and dental clinic ubiquitous applications).


[11_MSc_rosa]
Otávio Araújo Leitão ROSA. Test-driven maintenance: uma abordagem para a manutenção de sistemas legados. [Title in English: Test-driven Maintenance: an approach for the maintenance of legacy systems]. M.Sc. Diss. Port. Presentation: 08/04/11. 98 p. Advisor: Arndt von Staa.

Abstract: Test-Driven Development is a software development technique based on quick cycles that switch between wrinting test and implementing a solution that assures that tests do pass. Test-Driven Development has produced excellent result in various aspects of building new software systems. Incresead maintainability, improved design, reduced defect density, better documentation and increased code test coverage are reported as advantages that contribute to reducing the cost of development and, consequently, to increasing return on investiment. All these benefits have a contributed for Test-driven Development to become an increasingly relevant practice while developing software. When evaluating test driven development from the perspective of maintaining legacy systems, we face a clear mismatch when trying to adopt this technique. test-Driven Development is based on the premise that tests should be written before coding, but when working with legacy code we already have thousands of lines written and running. considering this context, we discuss in this dissertation a technique, which we call Test-Driven Maintenance, that is a result of adapting Test-Driven Development to the needs of maintaining legacy systems. We describe how we have performed the adaption that lead us to this new technique. Afterwards, we evaluate the efficacy of the technique applying it to a realistic project. To obtain realistic evaluation results, we have performed an empirical study while introducing the technique in a maintenance team working on a legacy system that is in constant evolution and use by an enterprise in Rio de Janeiro.


[11_MSc_vilela]
Paula de Castro Sonnenfeld VILELA. Classificação de sentimento para notícias sobre a Petrobras no mercado financeiro. [Title in English: Sentiment analysis for financial news about Petrobras company]. M.Sc. Diss. Port. Presentation: 01/07/11. 50 p. Advisor: Ruy Luiz Milidiú.

Abstract: A huge amount of information is available online, in particular regarding financial news. Current research indicate that stock news have a strong correlation to market variables such as trade volumes, volatility, stock prices and firm earnings. Here, we investigate a Sentiment Analysis problem for financial news. Our goal is to classify financial news as favorable or unfavorable to Petrobras, an oil and gas company with stocks in the Stock Exchange market. We explore Natural Language Processing techniques in a way to improve the sentiment classification accuracy of a classical bag of words approach. We filter on topic phrases for each Petrobras related news and build syntactic and stylistic input features. For sentiment classification, Support Vector Machines algorithm is used. Moreover we apply four feature selection methods and build a committee of SVM models. Additionally, we introduce Petronews, a Portuguese financial news annotated corpus about Petrobras. It is composed by a collection of one thousand and fifty online financial news from 06/02/2006 to 01/29/2010. Our experiments indicate that our method is 5.29% better than a standard bag-of-words approach, reaching 87.14%
accuracy rate for this domain.


[11_MSc_silveira]
Paulo da Silva SILVEIRA. Projeto e implementação de interfaces coletivas em um middleware orientado a componentes de software. [Title in English: Design and implementation of collective interfaces in a component-oriented middleware]. M.Sc. Diss. Port. Presentation: 08/04/11. 93 p. Advisor: Renato Fontoura de Gusmão Cerqueira.

Abstract: Traditionally, the development process of parallel systems emphasizes performance at the expense of better programming abstractions, which causes problems such as excessive code complexity and reduced software maintain-ability. New techniques have show expressive results in building parallel software, such as software components technologies. This work conducted a study of the mechanism of parallel communication between compnents known as Collective Interfaces. As part of this study, we performed an implementation of this mechanism in the SCS middleware, where two connectors were designed and implemented for parallel synchronization and communication. This implementation allowed us to analyze the requirements for the integration of Collective Interfaces in a component oriented middleware and to identify the challenges of implementing this mechanism in a language as C++, widely used in scientific applications.


[11_MSc_silva]
Paulo de Tarso Gomide Castro SILVA. A system for stock market forecasting and simulation. [Title in Portuguese: Um sistema para predição e simulação do mercado de capitais]. M.Sc. Diss. Port. Presentation: 25/03/11. 63 p. Advisor: Ruy Luiz Milidiú.

Abstract: he interest of both investors and researchers in stock market behavior forecasting has increased throughout the recent years. Despite the wide number of publications examining this problem, accurately predicting future stock trends and developing business strategies capable of turning good predictions into profits are still great challenges. This is partly due to the nonlinearity and noise inherent to the stock market data source, and partly because benchmarking systems to assess the forecasting quality are not publicly available. Here, we perform time series forecasting aiming to guide the investor both into Pairs Trading and buy and sell operations. Furthermore, we explore two different forecasting periodicities. First, an interday forecast, which considers only daily data and whose goal is predict values referring to the current day. And second, the intraday approach, which aims to predict values referring to each trading hour of the current day and also takes advantage of the intraday already known at prediction time. In both forecasting schemes, we use three regression tools as predictor algorithms, which are: partial Least Squares Regression, Support Vector Regression and Artificial Neural Networks. We also propose a trading system as a better way to assess the forecasting quality. In the experiments, we examine assets of the most traded companies in the BM&FBOVESPA Stock Exchange, the world's third largest and official Brazilian Stock Exchange. The results for the three predictors are presented and compared to four benchmarks, as well as to the optimal solution. The difference in the forecasting quality, when considering either the forecasting error metrics or the trading system metrics, is remarkable. If we consider just the mean absolute percentage error, the proposed predictors do not show a significant superiority. Nevertheless, when considering the trading system evaluation, it shows really outstanding results. The yield in some cases amounts to an annual return on investment of more than 300%.


[11_MSc_rodrigues]
Paulo Gallotti RODRIGUES. v-Glove: uma proposta de dispositivo de interação para aplicações imersivas de realidade virtual. [Title in English: v-Glove: proposing an interaction device for immersive virtual reality applications]. M.Sc. Diss. Port. Presentation: 01/04/11. 102 p. Advisor: Alberto Barbosa Raposo.

Abstract: Traditional interaction devices such as mouse and keyboard don't adapt to immersive applications, since their use in this kind of environment isn't ergonomic, because the user may be standing or in movement. Moreover, in the current interaction model for this kind of application (based on wands and 3D mice), the users have to change context every time they need to execute a non-immersive task, specially the symbolic input. These constant context changes from immersion to WIMP (Windows, Icons, Menus and Pointers) introduce a rupture in the user interaction with the application. The objective of this work is to explore the use possibilities of a device that maps a touch interface in a virtual reality immersive environment. We developed a glove for interaction in 3D virtual reality immersive environments (v-Glove), which has two main functionalities: tracking of the position of the user's forefinger in the space and the generation of a vibration in the fingertip when it reaches an area mapped in the interaction space. We performed quantitative and qualitative tests with users to evaluate v-Glove, comparing it with a 3D mouse used in immersive environments.


[11_MSc_souza]
Paulo Roberto França de SOUZA. Uma ferramenta para reconstrução da sequência de interações entre componentes de um sistema distribuído. [Title in English: A tool for rebuilding the sequence of interactions between components of a distributed system]. M.Sc. Diss. Port. Presentation: 04/04/11.  66 p. Advisor: Renato Fontoura de Gusmão Cerqueira.

Abstract: Distributed systems often present a runtime behavior different than what is expected by the programmer. Static analysis is not enough to understand the runtime behavior and to diagnoses errors. This difficulty is caused by the non-deterministic nature of distributed systems, because of their inherent characteristics such as concurrency, communication latency and partial failure. Therefore, it's necessary a better view of the interactions between the system's software components in order to understand its runtime behavior. In this work we present a tool that rebuilds the interactions among distributed components, presents a view of distributed threads and remote call sequences, and allows the analysis of causality relationships. Our tool also stores the interactions over time and correlates them to the system architecture and to performance data. The proposed tool helps the developer to better understand scenarios involving an unexpected behavior of the system and to restrict the scope of error analysis, making easier the search for a solution.


[11_PhD_mottajunior]
Paulo Rogério da MOTTA JUNIOR. Uma abstração para programação paralela: suporte para o desenvolvimento de aplicações. [Title in English: An abstraction for parallel programming: support for developing applications]. Ph.D. Thesis. Port. Presentation: 11/08/11. 107 p. Advisor: Noemi de La Rocque Rodriguez.

Abstract: The evolution of the field of programming traditionally trades performance for more powerful abstractions that are able to simplify the programmer’s work. It is possible to observe the effects of this evolution on the parallel programming area. Typically parallel programming focuses on high performance based on the procedural paradigm to achieve the highest possible throughput, but determining the point in which one should trade performance for more powerful abstractions remains an open problem. With the advent of new system level tools and libraries that deliver greater performance without programmer’s intervention, the belief that the application programmer should optimize communication code starts to be challenged. As the growing demand for large scale parallel solutions becomes noticeable, problems like code complexity, design and modeling power, maintainability, faster development, greater reliability and reuse, are expected to take part on the decision of which approach to use. In the present work, we investigate the use of higher-level abstraction layers that could provide many benefits for the parallel application developer. We argue that the use of interpreted languages may aid the abstraction of the processor architecture providing an opportunity to optimize the virtual machines without affecting the user’s application code.


[11_MSc_asti]
Pedro Larronda ASTI. Anotador morfossintático para o Português-Twitter. [Title in English: Morphosyntactic tagger for Portuguese-Twitter] M.Sc. Diss. Port. Presentation: 05/04/11. 49 p. Advisor: Ruy Luiz Milidiú.

Abstract: In this paper we present a language processor that solves the task of MORPHO-SYNTACTIC TAGGING of messages posted in Portuguese on Twitter. By analyzing the messages written by Brazilian on Twitter, it is easy to notice that new characters are introduced in the alphabet and also that new words are added to the language. Furthermore, we note that these messages are syntactically malformed. This precludes the use of existing Portuguese processors in these messages, nevertheless this problem can be solved by considering these messages as written in a new language, the Portuguese-Twitter. Both the alphabet and the vocabulary of such idiom contain features of Portuguese. However, the grammar is different. In order to build the processors for this new language, we have used a supervised learning technique known as ENTROPY GUIDED TRANSFORMATION LEARNING (ETL). Additionally, to train ETL processors, we have built an annotated corpus of messages in Portuguese-Twitter. We are not aware of any other taggers for the Morphosyntactic Portuguese-Twitter task, thus we have compared our tagger to the accuracy of state-of-art Morphosyntactic Annotation for Portuguese, which has accuracy around 96% depending on the tag set chosen. To assess the quality of the processor, we have used accuracy, which measures how many tokens were tagged correctly. our experimental results show an accuracy of 90,24% for the proposed MORPHO-SYSTATIC TAGGER. This corresponds to significant learning, since the initial baseline system has an accuracy of only 76,58%. This finding is consistent with the observed learning for the corresponding regular Portuguese taggers.


[11_MSc_moura]
Pedro Nuno de Souza MOURA. Integrando metaeurísticas com resolvedores MIP para o capacitated vehicle routing problem. [Title in English: Integrating metaheuristics with MIP solvers to the capacitated vehicle routing problem]. M.Sc. Diss. Port. Presentation: 26/08/11. 68 p. Advisor: Marcus Vinicius Soledade Poggi de Aragão.

Abstract: Since its inception, approaches to Combinatorial Optimization were polarized between exact and heuristic methods. Recently, however, strategies that combine both methods have been proposed for various problems, showing promising results. In this context, the concepts of ball and ellipsoidal neighborhood appear, which perform a search regarding one or more reference solutions. This work studies the application of such neighborhoods for the Capacitated Vehicle Routing Problem (CVRP), using the Robust Branchand-Cut-and-Price algorithm. Experiments were  made and its results were analyzed.


[11_MSc_riverasalas]
Percy Enrique RIVERA SALAS. StdTrip: an a priori design process for publishing linked data. [Title in Portuguese: StdTrip: um processo de projeto a priori para a publicação de "Linked Data"]. M.Sc. Diss. Port. Presentation: 01/04/11. 73 p. Advisor: Karin Koogan Breitman.

Abstract: Open Data is a new approach to promote interoperability of data in the Web. It consists in the publication of information produced, archived and distributed by organizations in formats that allow it to be shared, discovered, accessed and easily manipulated by third party consumers. This approach requires the triplication of datasets, i.e., the conversion of database schemata and their instances to a set of RDF triples. A key issue in this process is deciding how to represent database schema concepts in terms of RDF classes and properties. This is done by mapping database concepts to an RDF vocabulary, used as the base for generating the triples. The construction of this vocabulary is extremely important, because the more standards are reused, the easier it will be to interlink the result to other existing datasets. However, tools available today do not support reuse of standard vocabularies in the triplication process, but rather create new vocabularies. In this thesis, we present the StdTrip process that guide users in the triplication process, while promoting the reuse of standard, RDF vocabularies.
 

[11_MSc_pereira]
Rafael Silva PEREIRA. A split&merge architecture for distributed video processing in the Cloud. [Title in Portuguese: Uma arquitetura de split&merge para processamento distribuído de vídeo baseado em Cloud]. M.Sc. Diss. Eng. Presentation: 18/04/11. 76 p. Advisor: Karin Koogan Breitman.

Abstract: The Map Reduce approach, proposed by Dean& Ghemawat, is an efficient way for processing very large datasets using a computer cluster and, more recently, cloud infrastructures. Traditional Map Reduce implementations, however, provide neither the necessary flexibility (to choose among different encoding techniques in the mapping stage) nor control (to specify how to organize results in the reducing stage), required to process video files. The Split & Merge tool, proposed in this thesis, generalizes the Map Reduce paradigm, and provides an efficient solution that contemplates relevant aspects of intense processing video applications.


[11_MSc_marroquinmogrovejo]
Renato Javier Marroquín Mogrovejo. Experimental statistical analysis of MapReduce joins. [Title in Portuguese: Análise estadístico experimental de junções MapReduce]. M.Sc. Diss. Port. Presentation: 28/06/11. 89 p. Advisor: Edward Hermann Haeusler.

Abstract: There are different scalable data management solutions which can take advantage of cloud features making them more attractive for a deployment in such environments. One of the most critical operations in data processing is joining datasets, but this operation is most of the time the one that takes more time, and one of the hardest to optimize. In this work, we explore statistical methods in order to predict join queries execution times. In addition to that, join selectivity estimation is explored in a MapReduce environment in order to use it as another parameter in our model.


[11_PhD_espinha]
Rodrigo de Souza Lima ESPINHA. Suporte topológico em paralelo para malhas de elementos finitos em análises dinâmicas de fratura e fragmentação. [Title in English: Parallel topological support for finite element meshes in dynamic fracture and fragmentation analysIS]. Ph.D. Thesis. Port. Presentation: 05/04/11. 122 p. Advisor: Waldemar Celes Filho.

Abstract: Fracture propagation and fragmentation phenomena in solidis can be described by Cohesive Zone Models and simulated with the Finite Element Method. Among the computational approaches of recent interest for fracture representation in finite element meshes are those based on cohesive elements. In those approaches, fracture behavior is represented by cohesive elements inserted at the interfaces between volumetric (bulk) elements of the original mesh. Cohesive element models can be classified into intrinsic or extrinsic. Intrinsic models require pre-inserted cohesive elements at every volumetric interface in which fracture is allowed to happen. On the other hand, extrinsic models require that cohesive elements be adaptively inserted, wherever and wherever necessary. However, the traditional mesh representation (elements and nodes) is not sufficient for handling adaptative meshes, which makes an appropriate topological support necessary. In general, cohesive models of fracture also require a high level of mesh refinement near crack tips, such that accurate results can be achieved. this implies in memory and processor consumption that may be prohibitive for traditional workstations. Thus, parallel environments become important for the solution of fracture problems. However, due to the difficulties for thr parallelization of extrinsic models, the the existing approaches use intrinsic models or implement extrinsic simulations based on pre-inserted cohesive elements or cohesive elements represented as attributes of volumetric elements. In order to allow fracture and fragmentation simulations of large models in a simple and efficient way, this thesis proposes the ParTopS system, a parallel topological support for finite element meshes in dynamic fracture and fragmentation analyses. Specifically, a compact and efficient representation of distributed fracture meshes is presented. Cohesive elements are explicitly represented and treated as regular elements in the mesh. Based on the distributed mesh representation, we propose a scalable parallel algorithm for adaptive insertion of cohesive elements in both bidimensional and tridimensional meshes. Symmetrical topological operations are exploited in order to reduce communication among mesh partitions. The ParTopS system has been employed in order to parallelize existing serial extrinsic simulations. The scalability and correctness of the parallel topological support is demonstrated through computational experiments executed on a massively parallel environment. The achieved results show that ParTopS can be effectively applied in order to enable simulations of large models.


[11_PhD_correa]
Sand Luz CORREA. Detecção estatística de anomalias de desempenho em sistemas baseados em middleware. [Title in English: Statistical detection of performance anomalies in middleware-based systems]. Ph.D. Thesis. Port. Presentation: 01/04/11. 146 p. Advisor: Renato Fontoura de Gusmão Cerqueira.

Abstract: Middleware technologies have been widely adopted by the software industry to reduce the cost of developing computer systems. Nonetheless, predicting the performance of middleware-based applications is difficult due to specific implementation details of a middleware platform and a multitude of settings and services provided by middleware for different deployment scenarios. Thus, the performance management of middleware-based applications can be a non-trivial task. Autonomic computing is a new paradigm for building self-managed systems, i.e., systems that seek to operate with minimal human intervention. this work investigates the use of statistical approaches to building autonomic management solutions to control the performance of middleware-based applications. particulary, we investigate this issue from three perspectives. the first is related to the prediction of performance problems. We propose the use of classification techniques to derive performance models to assist the autonomic management of distributed applications. In this sense, different classes of models in statistical learning are assessed in both offline and online learning scenarios. The second perspective refers to the reduction of false alarms, seeking the development of reliable mechanisms that are resilient to transient failures of the classifiers. This work proposes an algorithm to augment the predictive power of statistical learning techniques by combining them with statistical tests for trend detection. Finally, thr third perspective is related to diagnosing the root causes of a performance problem. For this context, we also propose the use of statistical tests. The result presented in this thesis show that statistical approaches can contribute to the development of tools that are both effective, as well as efficient in characterizing the performance of middleware-based applications. Therefore, these approaches can contribute decisively to different perspectives of the problem.


[11_MSc_cerqueira]
Sérgio Luiz Ruivace CERQUEIRA. Comparação de projeto baseado em agentes e orientação a objetos na plataforma GeoRisc. [Title in English: Comparison of agent and object oriented projects using the GeoRisc platform]. M.Sc. Diss. Esp. Presentation: 05/04/11. 108 p. Advisor: Arndt von Staa.

Abstract: There are several software development technologies currently in the literature. Two such technologies are object orientation, which is consolidated, and agent orientation, which has been the subject of many studies and experiments. These studies indicate the agent orientation as very promising and an evolution of object orientation. However, there is only a few studies comparing these two techniques and these studies have been based on ideological and qualitative comparisons. This dissertation aims to develop and evaluate methods of systemic evaluation of two architecture for implementing systems. The two technologies presented were compared and determined whether the use of technology has brought benefits, disadvantages or was indifferent to the other. The comparison was performed based on taking a real problem; in other words, two implementations have been created that address the problem similary each using a technology. To develop this work, it was created a measurement plan based on the technique Goal Question Metric. The measurement plan was applied to both implementations and results were evaluated by defining the benefits of each technique. Finally was done a discussion about the use of the GQM model in a real project.


[11_MSc_buback]
Silvano Nogueira Buback. Utilizando aprendizado de máquina para construção de uma ferramenta de apoio a moderação de comentários. [Title in English: Using machine learning to build a tool that helps comments moderation]. M.Sc. Diss. Esp. Presentation: 01/09/11. 65 p. Advisors: Marco Antonio Casanova and Ruy Luiz Milidiú.

Abstract: One of the main changes brought by Web 2.0 is the increase of user participation in content generation mainly in social networks and comments in news and service sites. These comments are valuable to the sites because they bring feedback and motivate other people to participate and to spread the content. On the other hand these comments also bring some kind of abuse as bad words and spam. While for some sites their own community moderation is enough, for others this impropriate content may compromise its content. In order to help theses sites, a tool that uses machine learning techniques was built to mediate comments. As a test to compare results, two datasets captured from Globo.com were used: the first one with 657.405 comments posted through its site and the second with 451.209 messages captured from Twitter. Our experiments show that best result is achieved when comment learning is done according to the subject that is being commented.


[11_PhD_castro]
Thaís Helena Chaves de CASTRO. Sistematização da aprendizagem de programação em grupo. [Title in English: Systematic approach for group programming learning]. Ph.D. Thesis. Port. Presentation: 24/04/11. 152 p. Advisor: Hugo Fuks.

Abstract: The research reported here deals with devising structuring elements that may broaden intervention opportunities from the teacher in a context of group programming learning. Based on a set of case studies with freshmen in computing courses a systematization for practices, methods and technologies was developed producing an approach for supporting group programming based in three investigation paths: pedagogocal assumptions, CSCL environments and collaboration methods. The main learning rationale is Jean Piaget's Cognitive Development Theory, used alongside group programming techniques commonly applied in undergraduate introductory programming courses. Computational tools are used to monitor and intervene during learning process and in such context, CSCL environments encourage collaboration and regulate expected practices. In this thesis other technologies like languages for agent representation and patterning identification are also exploited for improving control and facilitate interventions. Finally, as collaborations method, it is proposed a Programming Progressive Learning Scheme that helps students to adopt collaborative practices when solving exercices and that can be formalized to be used with automated platforms.


[11_MSc_segura]
Vinicius Costa Villas Bôas SEGURA. UISKEI: sketching the user interface and its behavior. [Title in Portuguese: UISKEI: desenhando a interface e seu comportamento]. M.Sc. Diss. Port. Presentation: 30/03/11. 95 p. Advisor: Simone Diniz Junqueira Barbosa.

Abstract: During the early user interface design phase, different solutions should be explored and iteratively refined by the design team. In this rapidly evolving scenario, a tool that enables and facilitates changes is of great value. UISKEI takes the power of sketching, allowing the designer to convey his or her idea in a rough and more natural form of expression, and adds the power of computing, which makes manipulation and editing easier. More than an interface prototype drawing tool, UISKEI also features the definition of the prototype behaviour, going beyond navigation between user interface containers (e.g. windows, web pages, screen shots) and allowing to define changes to the state of user interface elements and widgets (enabling/disabling widgets, for example). This dissertation presents the main concepts underlying UISKEI and a study on how it compares to similar tools. The user interface drawing syage is detailed, explaining how the conversion of sketches to widgets is made by combining a sketch recognizer, which uses the Levenshtein distance as a similarity measure, and the interpretation of recognized sketches based on an evolution tree. Furthermore, it discusses the different solutions explored to address the issue of drawing an interaction, suggesting an innovative mind-map-like visualization approach that enables the user to express the event, conditions and actions of each interaction case still keeping the pen-based interaction paradigm in mind.