Theses and Dissertations
2012
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 2012. 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: 20/MARCH/2013
[12_MSc_pereira]
Adriano Brito PEREIRA.
Q-Learning Pessimista: um algoritmo para geração de bots de jogos em turnos. [Title
in English: Pessimistic Q-Learning: an algorithm to create bots for turn-based
games].
M.Sc. Diss. Port. Presentation: 26/07/12. 63 p. Advisor: Ruy Luiz Milidiú.
Abstract: This document presents a new algorithm for reinforcement learning
method,
Q-Learning Pessimistic. Our motivation is to resolve the problem of generating
bots able to play turn - based games and contribute to achieving better results
through this extension of the Q-learning algorithm. The Q-Learning Pessimistic
explores the flexibility of the calculations generated by the traditional
Q-learning without the use of force brute. To measure the quality of bot
generated, we consider quality as the sum of the potential to win and tie in a
game. Our fundamental purpose, is to generate bots with good quality for
different games. Thus, we can use this algorithm to families of turn - based
games. We developed a framework called Wisebots and conducted experiments with
some scenarios applied to the following traditional games TicTacToe, Connect - 4
and CardPoints. Comparing the quality of Pessimistic Q-Learning with the
traditional Q-learning, we observed gains to 100% in the TicTacToe, obtaining an
algorithm that never loses. Also observed in 35% gains Connect-4 and 27% in
CardPoints, increasing both the range of 60% to 80% for 90% to 100% of quality.
These results illustrate the potential for improvement with the use of
Q-learning Pessimistic, suggesting its application to various types of games.
[12_MSc_santos]
Adriano Medeiros dos
SANTOS.
Suporte a componentes compostos para o middleware SCS. [Title
in English: Composite component support for the SCS middleware].
M.Sc. Diss. Port. Presentation: 03/04/12. 99 p. Advisor: Renato Fontoura de
Gusmão Cerqueira.
Abstract: Several software component models support
composite components, such as Fractal, OpenCOM, Koala, Kobra and SaveCCM. A
composite component encapsulates components and provides introspection and
configuration mechanisms. Composite components are recognized as a good practice
because abstract complex structures and increase reuse. However, there is not a
comprehensive study about composite component mechanisms addressing issues like
initial composite configuration, mapping of subcomponents’services and
dependencies through composite component’s interfaces, introspection, component
sharing, and composite’s life cycle. In this work we perform a study about how
the Fractal and OpenCOM apply the composite component concept and we propose a
new model for composite components. As an experimental study we implemented a
new version of the SCScomponent system with support for composite components,
the SCS-Composite.We also present an evaluation of SCS-Composite through the
adaptation of a SCS based application.
[12_MSc_gurgel]
Alessandro Cavalcante GURGEL.
Blending and reusing rules for architectural degradation prevention. [Title
in Portuguese: Composição e reúso de regras para prevenção da degradação
arquitetural].
M.Sc. Diss. Eng. Presentation: 11/04/12. 87 p. Advisor: Alessandro Fabrício
Garcia.
Abstract: During the maintenance of software systems, their architecture
often degrades through processes of architectural erosion and drift. These
processes are often intertwined and, as a consequence, a given module in the
code becomes the locus of both erosion and drift symptoms. Architects should
elaborate strategies for detecting co-occurrences of both degradation symptoms.
Strategies for enabling the detection of these symptoms are based on design
rules. While the specification of design rules is time-consuming, they are often
similar across different software projects. In this context, the contribution of
this dissertation is threefold. First, it presents TamDera, an unified
domain-specific language for: (i) specifying rule-based strategies to detect
both erosion and drift symptoms, and (ii) promoting the hierarchical and
compositional reuse of design rules across multiple contexts. Second, a tool
implementation for supporting the language usage and rule enforcement is also
presented in this dissertation. Third, we evaluated the language in supporting
the description and reuse of design rules on five software projects. Our
evaluation revealed that architects could be benefited by using TamDera to blend
and reuse rules for detecting erosion and drift occurrences in multiple
scenarios.
[12_MSc_meracaraballo]
Alexander Arturo MERA CARABALLO.
Publishing
annotated multimedia Deep Web data. [Title
in Portuguese: Publicando anotações de dados multimídia advinda da Deep Web
].
M.Sc. Diss. Eng. Presentation: 03/04/12. 67 p. Advisor: Marco Antonio
Casanova.
Abstract: In recent years, we witnessed a huge growth of multimedia data on
the Web. New lower-cost technologies and greater bandwidth allowed the Web to
evolve into a multimedia format. However, the lack of tools that can make
multimedia format easily accessible on the Web led us to a non-searchable and
non-indexable data of the Web, also known as Deep Web. In line with these
observations, this dissertation addresses the problem of how to publish audio
and video content on the Web. We present a tool and a novel approach that
facilitates the indexing and retrieval of the objects with the help of
traditional search engines. The tool automatically generates static Web pages
that describe the content of the objects and organize this content to facilitate
locating segments of the audio or video which correspond to the descriptions.
The static Web pages can be translated to others languages to reach other user
populations. An annotation process is also performed to embed machine-readable
data into the Web pages. The dissertation also presents an in-depth experiment,
publishing learning objects based on audio and video, to assess the efficacy of
the technique.
[12_PhD_costa]
Andrew Diniz da COSTA.
Automação do processo de gerência
de teste de software. [Title in English: Automation of the management process of
the test of software]. Ph.D. Thesis. Port. Presentation:10/08/2012. 171 p.
Advisor:Carlos José Pereira de Lucena.
Abstract:Testing software systems has become a critical activity of software
development over time. Especially when the development of complex systems,which
are able to self-adapt theirs behaviors when necessary, is becoming extremely
common. In this context, the multi-agent system (MAS) paradigm has been used
especially when distributed, autonomous and pro-active entities are represented.
Aiming to control the creation, maintenance and execution of tests on these
systems, it is necessary to adopt a management process that considers the
activities of planning, design, execution and monitoring of the tests. In order
to help this management, the thesis presents the Java self-Adaptive Agent
Framework for Self-Test (JAAF+T), that aims
to allow the creation of selfadaptive agents that perform self-tests. Since
several data are used to help the management of the tests, documenting them
helps to understand how to evolve and execute them. Based on this concern, the
thesis offers a new modeling language called UML Testing Profile for
Coordination (UTP-C), profile of the UML that allows the modeling of these data.
At last, but not least important, tools, which were created to automate the
management of the tests executed for selfadaptive agents, are presented to
generate useful artifacts used for instances of the JAAF+T based on UTP-C models.
[12_PhD_santosneto]
Baldoíno Fonseca dos SANTOS NETO.
Uma abordagem deôntica para o desenvolvimento de agentes normativos autônomos.
[Title in English: A deontic apporach to develop autonomous normative agents]. Ph.D. Thesis.
Port. Presentation: 30/03/2012. 169 p. Advisor: Carlos José Pereira de Lucena.
Abstract: Social norms have become one of the most promising approaches
toward ensuring a desirable social outcome in multi-agent systems. In these
systems, autonomous and heterogeneous agents work toward common or different
goals. Norms regulate the behaviour of these agents by defining obligations and
prohibitions, and by creating rewards and penalties to encourage the agents to
behave so as to meet these norms. Although the use of norms is a promising
mechanism to regulate the agents’ behavior, the development of agents able to
join an agents society, to adopt the norms and to work with them autonomously
has shown to be significantly more challenging than traditional agents. The
reason for this difficulty is the lack of agent models that allow developing of
autonomous agents able to cope with issues related to norms. In this thesis,
we introduce an approach to develop agents able to deal with normative questions
in an autonomous way. The approach extends the Belief-Desire- Intention model
including functions that assist the agent in adopting new norms, in verifying
norm activation, deactivation, fulfillment and violation, in selecting the norms
to be fulfilled or violated by the agent, in identifying and overcoming
conflicts among norms, in generating new goals from norms and in selecting goals,
plans and intentions while taking into account the system norms. In addition,
the approach provides mechanisms that are necessary to implement agents designed
from the proposed architectural model. Two scenarios are used in order to
exemplify the proposed approach. The first is related to the planning of rescue
missions regulated by norms. The second is related to the support to software
development regulated by norms. Finally, experiments that demonstrate the
importance of developing agent by using the proposed architectural model, both
in the agent and society point of view, are presented.
[12_MSc_araujo]
Bruno Baere Pederassi Lomba de ARAÚJO.
Um estudo sobre adaptatividade dinâmica de dificuldade em jogos. [Title in English: A study on
dynamic difficulty adaptivity in games] M.Sc.Diss. Port. Presentation: 06/09/12.
107 p. Advisor: Bruno Feijó.
Abstract: The areas of player modeling and adaptive systems are relevant to
games by allowing the game to adapt itself to the player’s skill level, trying
to offer a continuous challenge to the player, relative to its evolution in a
way the player feels motivated to keep playing by the challenge proportional to
his performance.
The present work is a study on player modeling and dynamic difficulty adjustment,
collecting data about games that use such concepts. An adaptive system based on
player modeling is developed and tested with players to analyze the
effectiveness and relevance of using such systems for maintaining the interest
of the player, from the perspectives of the flow theory and the core elements of
the gaming experience (CEGE) model.
[12_MSc_valentim]
Caio Dias VALENTIM.
Estruturas de dados para séries
temporais. [Title in English: Data structures for time series]. M.Sc. Diss. Port.
Presentation: 31/07/12. 57 p. Advisor: Eduardo Sany Laber.
Abstract: bstract: Time series are important tools for the anaylsis of
events that occur in diferent fields of human knowledge such as medicine,
physics, meteorology and finance. A common task in analysing time series is to
try to find events that happen infrequently as these events usually reect facts
of interest about the domain of the series. In this study, we develop techniques
for the detection of rare events in time series. Technically,
a time series A = (a1; a2; : : : ; an) is a sequence of real values indexed by
integer numbers from 1 to n. Given an integer t and a real number d, we say that
a pair of time indexes i and j is a (t; d)-event in A, if and only if 0 < j ?? i
_ t and aj ?? ai _ d. In this case, i is said to be the beginning of the event
and j is its end. The parameters t and d control, respectively, the time
window in which the event can occur and magnitude of the variation in the
series. Thus, we focus on two types of queries related to the (t; d)-events,
which are:
-What are the (t; d)-events in a series A? - What are the indexes in the series
A which are the beginning of at least one (t; d)-event? Throughout this study we
discuss, from both theoretical and practical points of view, several data
structures and algorithms to answer the two queries mentioned above.
[12_PhD_nunes]*
Camila Patricia Bazilio NUNES. History-sensitive recovery
of features in code of evolving program families. [Title in Portuguese:
Recuperação sensível a história de características no código de famílias de
programas evolutivas ]. Ph.D. Thesis.Eng. Presentation: 19/10/12. 155 p. Advisors:
Carlos José Pereira de Lucena and Alessandro Fabricio Garcia.
Abstract: A program family might degenerate due to unplanned
changes in its implementation, thus hindering the maintenance of family members.
This degeneration is often induced by feature code of the program family that is
changed individually in each member without considering other family members. In
extreme cases, the program family code is fully or partially replicated and
individually changed across several evolving members. Hence, as a family evolves
over time, it might no longer be possible to identify and classify the
implementation elements realizing common and variable features. One of the
imminent activities to address these problems is the history-sensitive recovery
of program family’s features. This recovery process encompasses the historical
analysis of each family member in order to identify and classify the
implementation elements (i.e. methods, attributes) according to their
variability nature. Existing work fails to analyse the evolution of the family
members with the goal of recovering features’ implementation elements.
Additionally, existing techniques for feature analysis are not effective as they
only take into consideration the history of a single member product. In summary,
the contributions of this thesis are threefold: (i) a catalogue of mapping
mismatches to guide software engineers in promoting the correctness and
completeness of their feature mappings. This catalogue is useful to ensure a
better effectiveness of the recovery process during the mapping analysis; (ii) a
suite of five heuristics for the automatic expansion of feature mappings
throughout the program family history. Those heuristics rely on both the
multi-dimensional historical analysis of program families and the catalogue of
mapping mismatches; and (iii) a suite of history-sensitive heuristics for
classifying the implementation elements realizing each family feature according
to their variability degree.
[12_MSc_rapello]*
Demetrius Costa RAPELLO. Sistema de recomendação de
segundo nível para suporte à produção de matérias jornalísticas. [Title in
English: Second Level recomendation system to support news editing]. M.Sc. Diss.
Port. Presentation: 15/03/12. 72 p. Advisor: Marco Antonio Casanova.
Abstract: Recommendation systems are widely used by major Web portals due to
the increase in the volume of data available on the Web. Such systems are
basically used to suggest information relevant to their users. This dissertation
presents a second-level recommendation system, which aims at assisting the team
of journalists of a news Web portal in the process of recommending related news
for the users of the Web portal. The system is called second level since it
creates recommendations to the journalists Who, in turn, generate
recommendations to the users. The system follows a model based on features
extracted from the text itself. The extracted features permit creating queries
against a news database. The query result is a list of candidate news, sorted by
score and date of publication, which the news editor manually processes to
generate the final list of related news.
[12_MSc_marx]
Edgard Luiz MARX.
Babel: um framework extensível para a publicação de RDF de várias fontes de
dados utilizando templates. [Title
in English: Bebel: an extensible framerwork for easy RDF publication from
multiple data sources using templates].
M.Sc. Diss. Port. Presentation: 11/02/12. 107 p. Advisor: Karin Koogan
Breitman.
Abstract: The vast majority of data on the Web today is not
Semantic Web ready. To facilitate and promote the conversion of data, stored in
relational databases and spreadsheets in particular, we introduce the Babel
approach. Differently from existing approaches, notably RDBtoRDF, Babel outputs
data in a wider range of formats, that include OWL, RDFa, RSS and (X)HTML, in
addition to RDF. The main contribution of Babel, however, is its ease of use.
Babel smoothes the learning curve by altogether eliminating the need of getting
acquainted with complex mapping techniques, that are substituted by the use of
templates.
[12_MSc_dallafavera]
Eduardo Ceretta DALLA FAVERA.
Oclusão de ambiente usando traçado de cone com voxelização da cena. [Title
in English: Ambient occlusion using cone tracing with scene voxelization].
M.Sc. Diss. Port. Presentation: 04/04/12. 64 p. Advisors: Waldemar Celes
Filho.
Abstract: Ambient occlusion is a low-cost technique to simulate indirect
ambient illumination in a realistic way. In this dissertation, we propose an
ambient occlusion method that achieves good quality results in real time. Using
an efficient voxelization algorithm, we create a volumetric description of the
geometry in a regular grid. During the scene rendering, the hemisphere around
each visible point is sampled by several cones, each one corresponding to the
trace of multiple rays. The volume of each cone is sampled by a serie of spheres,
which are used to compute the cone’s volume obstructed by the scene geometry.
The occlusion caused by the geometry at each visible point is estimated by
accumulating each sphere result in the cones and each cone result in the
hemisphere. We found this approach to be adequate, since the intersection of
each sphere with the voxelized scene can be performed in a very efficient manner.
[12_MSc_araujo]
Eduardo
Cruz ARAÚJO.
Projetando aplicações para TVDI através de storyboards interativos. [Title
in English: Designing applications for IDTV with interactive storyboards].
M.Sc. Diss. Port. Presentation: 04/04/12. 63 p. Advisors: Luiz Fernando Gomes
Soares.
Abstract: The adoption of the middleware Ginga as standard for
interactive digital TV is becoming increasingly established in Brazil and in
much of Latin America. However, the possibilities created by the middleware are
still not well explored by content producers. Much of the currently interactive
applications broadcasted has been restricted to just applications with the
additional content with little or no relation with the main audiovisual content.
Among other factors, when the author is interested in the design of interactive
content from the design stage of audiovisual content, would favor the appearance
of truly interactive programs, in which, for example, the viewer could decide
the destiny of a particular character or the end of a film according to their
preferences. This work presents a tool that allows authors of interactive
applications for digital TV to design your application using interactive
storyboards. Thus, several alternative interaction and layout can be tested even
before the production of all audiovisual content, and much interactivity as the
main video content can be thought together. Thereby, much richer interactive
programs can be made, in which, for example, the viewer may decide the fate of a
specific character or the progress of a movie. Moreover, since all final
audiovisual content were produced, it can also be incorporated into the tool in
order to make the final adjustment and generate the final application, without
any knowledge of programming languages.
[12_MSc_barbosa]
Eiji
Adachi Medeiros BARBOSA.
Sistema de recomendação para código de tratamento de
exceções. [Title
in English: Recommender system for exception handling code].
M.Sc. Diss. Port. Presentation: 30/03/12. 125 p. Advisor: Alessandro Fabricio
Garcia.
Abstract: Exception handling mechanisms are models commonly used
in programming languages to structure the exceptional execution flow of software
modules. These mechanisms enable the detection and signaling of exception
occurrences, allowing these modules to take corrective actions to recover a
module to a correct state. Recent studies show that recurrent failures and
performance issues in software applications are
related to the poor structuring of exception handling code. The preliminary
studies performed in the context of this dissertation not only confirm these
results, but also show other interesting facts that occur during software
evolution. In particular, it was realized that it is often very common to
observe ineffective exception handling actions in the first releases of a
system. It was also realized that little (or no) effort was allocated to improve
the quality of these action during the software evolution. Recent studies that
aim aiding software developers to deal with exception handling elements fail in
assisting them in a crucial task: implementing well-structured exception
handling code since the first release of
a system. In this context, this dissertation proposes a recommender system able
to provide code fragments implementing exception handling code. The proposed
strategy aims to assist developers in implementing their own code in their
applications. The system extracts structural information from the code under
development to search for examples in a local repository in order to recommend
code examples that share similar structural context with the code that the
developer is working on. The system implemented in the context of this
dissertation recommends code examples implemented in Java.
[12_PhD_cirilo]
Elder José Reioli CIRILO.
Supporting heterogeneous configuration knowledge of
software product lines. [Title in Portuguese: Suportando conhecimento de
configuração heterogêneo de linha de software]. Ph.D. Thesis. Eng. Presentation:
10/04/11. 104 p. Advisor: Carlos José Pereira de Lucena.
Abstract:
Customer-specific software can provide a competitive advantage compared to
general-purpose software. A software product line (SPL) aims at tailoring
customer-specific products from a set of reusable features. It is well known
that systematic reuse of features potentially leads to significant gains such as
faster production with low costs and high quality. In practice, building
enterprise software product lines might become a laborious and error-prone task
when we take into account the use of current implementation techniques.
Successful development of modern enterprise software product lines requires the
convergence of multiple views (e.g., domain experts, interface designers, and
developers), which are usually managed via domain-specific concepts. In this
case, each participant of the development process has a particular working
language to solve specific problems to its expertise. The challenge of
integrating these differences into the engineering process, avoiding the
potential cacophony of multiple different languages, is the heterogeneous
configuration knowledge
problem. In this thesis, we examine the current shortcomings on the
specification of heterogeneous configuration knowledge and as a solution we
propose the notion of Domain Knowledge Modeling Languages (DKMLs). The purpose
of DKMLs is to put in evidence the domain concepts and their programming
interfaces, which help to reduce source code obfuscation and increase feature
comprehension. Moreover, evidencing the domain-specific concepts, we are also
able to prevent inconsistencies on products by detecting errors in the entire
software product line. Another result of this research is the GenArch+, an
extensible tool that supports the flexible incorporation of DKMLs in traditional
software product line engineering. We emphasize the benefits of GenArch+,
including simplicity, expressiveness, and being framework independent. Finally,
we illustrate and evaluate the use of DKMLs in three different product lines.
[12_PhD_fernandes]
Eraldo Luis Rezende FERNANDES.
Entropy
guided feature generation for structure learning [Title in Portuguese:
Geração de atributos guiada por entropia para
aprendizado de estruturas]. Ph.D. Thesis. Eng. Presentation: 06/09/12. 93 p. Advisor:Ruy Luiz Milidiú.
Abstract: Structure learning consists in
learning a mapping from inputs to structured outputs by means of a sample of
correct input-output pairs. Many important problems fit into this setting.
Natural language processing provides several tasks that can be formulated and
solved as structure learning problems. Dependency parsing, for instance,
involves the prediction of a tree underlying a sentence. Feature generation is
an important subtask of structure learning which, usually, is partially solved
by a domain expert that builds complex discriminative feature templates by
conjoining the available basic features. This is a limited and expensive way to
generate features and is recognized as a modeling bottleneck. In this work, we
propose an automatic feature generation method for structure learning problems.
This method is entropy guided since it generates complex features based on the
conditional entropy of local output variables given the available input features.
We experimentally compare the proposed method with two important alternative
feature generation methods, namely manual template generation and polynomial
kernel methods. Our experimental findings indicate that the proposed method is
more attractive than both alternatives. It is much cheaper than manual templates
and computationally faster than kernel methods. Additionally, it is simpler to
control its generalization performance than with kernel methods. We evaluate our
method on nine datasets involving five natural language processing tasks and
four languages. The resulting systems present state-of-the-art comparable
performances and, particularly on part-of-speech tagging, text chunking,
quotation extraction and coreference resolution, remarkably achieve the best
known performances on different languages like Arabic, Chinese, English, and
Portuguese. Furthermore, our coreference resolution systems achieve the very
first place on the Conference on
Computational Natural Language Learning 2012 Shared Task. The competing systems
were ranked by the mean score over three languages: Arabic, Chinese and English.
Our approach obtained the best performances among
all competitors for all the three languages. Our feature generation method
naturally extends the general structure learning framework and is not restricted
to natural language processing tasks.
[12_MSc_diasjunior]
Eugênio
Pacelli Ferreira DIAS JUNIOR.
Aprendizado por reforço sobre o problema de
revisitação de páginas web. [Title
in English: Using reinforcement learning on web pages revisitation problem].
M.Sc. Diss. Port. Presentation: 19/01/12. 73 p. Advisor: Eduardo Sany Laber.
Abstract: In the Internet, the information we desire is usually spread over
different locations. For some applications, it is necessary to maintain local
copies of this information. Keeping consistency as well as freshness of a data
base, or more specifically a set of internet web pages, is a task systematically
studied. An approach to this problem is the use of reinforcement learning
techniques, using dynamic programming and stochastic
analysis to obtain a good rescheduling policy for the web pages copies. This
work is proposed to validate the use of reinforcement learning techniques over
this problem, as well as finding features of the problem useful to model the
developed solution.
[12_MSc_flores]
Evandro Oliveira das FLORES.
Uma análise de práticas na aplicação de Scrum em projetos de grande porte. [Title
in English:An Analysis of Practices in Applying SCRUM on large projects]. M.Sc.
Diss. Port. Presentation: 05/03/12. 79 p. Advisor: Arndt Von Staa
Abstract: In today's writings, there are many examples of using Scrum in
small teams and projects, leaving a knowledge gap about the possibility of
applying Scrum in big teams and large projects. This work aims at presenting
case studies in known companies where Scrum has been applied in large projects,
emphasizing the roadblocks found throughout the process, and the solutions
adopted, highlighting the practices that lead the projects to success.
[12_MSc_silva]
Fabricio Cardoso da SILVA.
Detalhamento de superfcie utilizando tesselação em
hardware. [Title in English: Surface detailing using hardware tessellation].
M.Sc. Diss. Port. Presentation: 29/06/12. 62 p.Advisor: Alberto Barbosa Raposo.
Abstract: Bump mapping techniques are widely used to simulate structural
details of tridimensional surfaces in order to improve visual quality and
compensate for the low geometric detailing generally applied to models sent to
the GPU due to performance issues. Recent advances in the rendering pipeline
enabled the massive generation of vertex data in the graphics hardware by means
of the tessellation feature, providing developers with a powerful tool to
control the meshes’ level of details. The present work proposes a technique for
geometric
detailing of models using hardware tessellation, both based on pre-computed
depth maps and on displacement data generated entirely on the GPU through
procedural textures techniques. Performance and visual quality analysis
demonstrates the advantages of the proposed method in relation to an image-based
technique commonly used in videogames for enhancing the visual quality of the
environments.
[12_PhD_benjamimfilho]
Francisco
BENJAMIM FILHO.
Classificação de páginas Web por aprendizagem de múltiplas
categorias latentes. [Title
in English: Ranking of Web pages by learning multiple latent categories].
Ph.D. Thesis. Port. Presentation: 27/03/12. 76 p. Advisor: Ruy Luiz Milidiú.
Abstract: The rapid growth and generalized accessibility of the World Wide
Web (WWW) have led to an increase in research in the field of the information
retrieval for Web pages. The WWW is an immense and prodigious environment in
which Web pages resemble a huge community of elements. These elements are
connected via hyperlinks on the basis of similarity between the content of the
pages, the popularity of a given page, the extent to which the information
provided is authoritative in relation to a given field etc. In fact, when the
author of a Web page links it to another, s/he is acknowledging the importance
of the linked page to his/her information. As such the hyperlink structure of
the WWW significantly improves research performance beyond the use of simple
text distribution statistics. To this effect, the HITS approach introduces two
basic categories of Web pages, hubs and authorities which uncover certain hidden
semantic information using the hyperlink structure. In 2005, we made a first
extension of HITS, called Extended Hyperlink Induced Topic Search (XHITS), which
inserted two new categories of Web pages, which are novelties and portals. In
this thesis, we revised the XHITS, transforming it into a generalization of
HITS, broadening the model from two categories to various and presenting an
efficient machine learning algorithm to calibrate the proposed model using
multiple latent categories. The findings we set out here indicate that the new
learning approach provides a more precise XHITS model. It is important to note,
in closing, that experiments with the ClueWeb09 25TB collection of Web pages,
downloaded in 2009, demonstrated that the XHITS is capable of significantly
improving Web research efficiency and producing results comparable to those of
the TREC 2009/2010 Web Track. [not available for download yet]
[12_MSc_sousa]
Henrique Prado SOUSA.
Integrando modelagem intencional à modelagem de processos. [Title
in English: Integrating intentional modeling to process modeling].
M.Sc. Diss. Port. Presentation: 17/02/12. 138 p. Advisor: Julio Cesar Sampaio
do Prado Leite.
Abstract: The business processes modeling is used by
companies who wish to document details of the execution flow of their processes,
resulting in a document rich in details about the business. This artifact is
also used by the Software Engineering for system requirements elicitation. The
intentional modeling is focused on objectives - defined as goals and softgoals -
and registers the strategies that may be followed by an actor
in a way to better meet their needs, mapping the tasks and resources needs, in
addition, it also addresses the dependencies between actors. It is important
that business processes models are aligned to the objectives of the organization
in order to provide reliable information source that generates consequently
requirements aligned to business. Several tools are available on the market in
order to support the business processes
and organizational objectives modeling, however, it’s possible to realize that
the available solutions are still incomplete when it comes to the integration of
process models and goals models and ways to check the alignment between
organizational goals and processes using the models. In the organizational
architecture, business processes and goals are intrinsically interdependent,
however, the current modeling languages treat process and goals in a misaligned
way, since there are deficiencies in the integration between the modeling layer
of objectives and processes. Thus, the use of the available tools that supports
these language and methods greatly complicates the task of identify if the
processes used to generate products and services truly achieve the
organizational goals as well as the impact of the changes in the goals would
cause in business processes. In this paper we integrated a goal modeling
language to a business processes modeling language and proved the elements and
methods needed to expand the capacity of analysis of the alignment between the
business processes and the organizational strategies.
[12_PhD_nunes]*
Ingrid Oliveira de NUNES. User-centric preference-based decision making. [Title
in Portuguese: Tomada de decisão baseada em preferências e centrada no usuário].
Ph.D. Thesis. Eng. Presentation: 20/09/12. 298 p. Advisor: Carlos José Pereira
de Lucena.
Abstract: Choosing from a set of available options often requires resolution
of trade-offs but it can be unfeasible for humans to carefully evaluate each
option of a large set due to the required time and cognitive effort.
Consequently, they are often unsatisfied with their choices. Software systems
can support human decision
making or even automate this process, but there are many challenges associated
with the provision of such support. In this thesis we deal in particular with
three of them: (i) how to represent user preferences; (ii) how to reason about
preferences and make decisions; and (iii) how to justify such decisions.
Different approaches have been proposed for representing and reasoning about
qualitative preferences, but they address a restricted set of preference types,
and therefore are not able to process preferences provided by users in many
realistic scenarios. This thesis provides three main contributions. First, we
introduce a new preference metamodel founded on a study of how humans express
preferences, allowing the representation of high-level preferences.
Second, we propose an automated decision making technique, which chooses an
option from a set available based on preferences expressed in a language based
on our metamodel, exploiting natural-language terms. Our technique goes beyond
the provided preferences to make a decision with the incorporation of psychology
principles, which concern how humans make decisions, as the provided preferences
are typically not enough to resolve trade-offs among available options.Third, we
present an explanation generation technique, which uses models built by our
decision making technique to justify choices, and follows guidelines and
patterns that we derived from a study of choice explanation. A user study was
performed to evaluate our approach, which shows that (i) our preference language
is adequate for users to express their preferences, (ii) our decision making
technique makes choices that users consider as having good quality, and (iii)
the provided explanations allow users to understand why the choice was made and
improves the confidence in the decision.
[12_PhD_souza]
Jairo Francisco de SOUZA.
Uma abordagem heurística uni-objetivo para calibragem
em meta-alinhadores de ontologias. [Title in English: A single goal
heuristic approach to solve tunning in ontology meta-matching].
Ph.D. Thesis. Port. Presentation: 28/06/12.
105p. Advisor: Carlos José Pereira de Lucena
Abstract: Ontlogies are seen
as a solution for many applications such as database integration, e-commerce,
web services and social networks. However, the use of distintic ontologies does
not solve the semantic interoperability problem among models. Matcher selection
and self-configuring are challenges from the ontology matching field. this work
aims to treat the ontology meta-matching problem using heuristic methods to fast
tune a set of matchers. To evalutate this proposal, a well-kown benchmark for
ontology matching is used. The compasion among meta-matching approaches includes
the benchmark evaluation metrics and qualitative metrics for meta-matching
proposed in the literature.
[12_MSc_silveira]
Jaumir Valença da SILVEIRA JUNIOR.
Dealing with device data overflow in the
cloud. [Title
in Portuguese: Utilizando a nuvem para lidar com a sobrecarga de dados em
dispositivos fixos e móveis].
M.Sc. Diss. Port. Presentation: 29/03/12. 64 p. Advisor: Karin Koogan
Breitman.
Abstract: Cloud computing is rapidly becoming an important
platform for research in Software Engineering. Despite the vibe and huge
literature on commercial Cloud environments, there is, however, little research
on how to capture, model, design and implement new software applications that
can make intelligent use of the Cloud. In this paper we propose a new
abstraction that explores a fundamental aspect of Cloud systems – data
elasticity. The Container Database (CDB) abstraction provides a Cloud-based
solution for scenarios where device local storage is not sufficient for
manipulating data. To demonstrate the viability of the proposed approach we
present an implementation of the CDB abstraction as an Object-Oriented API
designed to work on Windows 7 and Windows Mobile Phone 7 Operation Systems.
[12_MSc_silveira]
Joner Martins Veiga DUARTE JUNIOR.
Um framework para o uso colaborativo de
dispositivos móveis como ferramenta de controle remoto de aplicações
científicas. [Title in English: A framework for collaborative use of mobile
devices for remote control of scientific applications]. M.Sc. Diss. Port.
Presentation: 12/09/12. 51 p. Advisor: Waldemar Celes Filho.
Abstract: Nowadays, mobile devices have become very popular bringing new
ways of interaction with their touch-based interface. Scientific visualization
applications have a great potential to take advantage of this new kind of
interaction, but the processing capabilities of mobile devices are still not
enough to render or process the amount of data this type of application requires.
We propose a framework, working as a client-server model, which allows the use
of mobile devices to collaboratively visualize and manipulate scientific
visualization applications. In the server role, the scientific application uses
a library to compress and send rendered images to clients and also to process
received events. In the client role, there is a multiplatform application
(iOS/Android) running on mobile devices, which interpret touch gestures and show
the images received through Wi-Fi network. The same application is able to
connect in any server, since it builds its interface from
a description in Lua language supplied by the server. Lastly, we evaluate the
proposed framework with two industrial applications: Geresim e 3DReplay.
[12_MSc_talaveraherrera]
José Eduardo TALAVERA HERRERA.
Arquitetura para recomendação de fontes de dados RDF. [Title in English: An architecture for RDF data sources recommendation.]
M.Sc. Port. Presentation: 01/10/12. 112 p. Advisor: Karin Breitman.
Abstract: In the Web publishing process of data it is recommended to link
the data from different sources using similar resources that describe a domain
in common. However, the growing number of published data sets on the Web have
made the data discovery and data selection tasks become increasingly complex.
Moreover, the distributed and interconnected nature of the data causes the
understanding and analysis to become too prolonged. In this context, this work
aims to provide a Web architecture for identifying RDF data sources with the
goal of improving the publishing, interconnection, and data exploration
processes within the Linked Open Data. Our approach utilizes the MapReduce
computing model on top of the cloud computing paradigm. In this manner, we are
able to make parallel keyword searches over existing semantic data indexes
available on the web. This will allow to identify candidate sources to link the
data. Through this approach, it was possible to integrate different semantic web
tools and relevant data sources in a search process, and also to relate topics
of interest defined by the user. In order to achieve our objectives it was
necessary to index and analyze text to improve the search of resources in the
Linked Open Data. To show the effectiveness of our approach we developed a case
study using a subset of data from a source in the Linked Open Data through its
SPARQL endpoint service. The results of our work reveal that the generation and
usage of data source's statistics do make a great difference within the search
process. These statistics help the user within the choosing individuals process.
Furthermore, a specialized keyword extraction process is run for each individual
in order to create different search processes using the semantic index. We show
the scalability of our RDF recommendation process by sampling several
individuals.
[12_MSc_salomao]
Khalid Fernandes Braga SALOMÃO.
Um
framework de jogos sérios para mercado de
informações com aplicação em ética empresarial.
[Title
in English: A serious game framework for information market with application in
business ethics]. M.Sc. Diss. Port. Presentation: 23/03/12. 103 p. Advisor: Bruno Feijó.
Abstract: This
dissertation focuses on implementing a framework to streamline the development
of game using the model of Informational Markets. Basically it consists in using
common sense and participants wisdom in a speculative market structure whose
objects traded are, in fact, the possible outcomes of the event in question. The
Informational Market is implemented in the form of stock exchange, in which
participants invest money in buying shares, which reflect their beliefs in a
particular outcome. This framework focuses on serious games that aim at the
application, contextualization and enrichment of knowledge of the player. The
framework allows configuration and running games with a wide range of topics. In
the context of this dissertation, the framework is used to implement an
Informational Market about a broad and general concept, which is ethics in
organizations. A guiding tool when defining ethical conduct in organizations by
providing an instrument that involves practical information of the perceptions
of ordinary people about the image of companies and about what they perceive as
the most important elements in the sense of ethical actions. The serious game is
implemented as a website with ASP.NET technology, using the framework to run the
simulation.
[12_PhD_oliveira]*
Kleinner Silva Farias de OLIVEIRA.
Empirical evaluation of effort on composing design models. [Title
in Portuguese: Avaliação empírica de esforço em composição de modelos de projeto].
Ph.D. Thesis. Eng. Presentation: 21/03/12. 282 p. Advisor: Alessandro Fabricio
Garcia and Carlos José Pereira de Lucena.
Abstract: Model composition
plays a central role in many software engineering activities such as evolving
models to add new features and reconciling conflicting design models developed
in parallel by different development teams. As model composition is usually an
error-prone and effort-consuming task, its potential benefits, such as gains in
productivity can be compromised. However, there is no empirical knowledge
nowadays about the effort required to compose design models. Only feedbacks of
model composition evangelists are available, and they often diverge.
Consequently, developers are unable to conduct any cost-effectiveness analysis
as well as identify, predict, or reduce composition effort. The inability of
evaluating composition effort is due to three key problems. First, the current
evaluation frameworks do not consider fundamental concepts in model composition
such as conflicts and inconsistencies. Second, researchers and developers do not
know what factors can influence the composition effort in practice. Third,
practical knowledge about how such influential factors may affect the developers’
effort is severely lacking. In this context, the contributions of this thesis
are threefold: (i) a quality model for supporting the evaluation of model
composition effort, (ii) practical knowledge, derived from a family of
quantitative and qualitative empirical studies, about model composition effort
and its influential factors, and (iii) insight about how to evaluate model
composition efforts and tame the side effects of such influential factors.
[12_MSc_guimaraes]
Leandro Fernandes GUIMARÃES.
Um framework para desenvolvimento de agentes autoadaptativos em dispositivos móveis. [Title
in English: A framework for developing self-adaptive agents for mobile devices]. M.Sc. Diss. Port. Presentation: 03/03/12. 95 p. Advisor: Carlos José Pereira de Lucena.
Abstract:
Technological progress in mobile computing and the Internet promotes an
achievement of the futuristic vision in which users have access to data and
services anytime and anywhere. However, this vision realization brings important
issues to application development, because more attention is required to the
communication between mobile devices and web services, especially regarding
processing and data storage constraints, diversity standards and availability of
services. This dissertation explores aspects of software engineering as
autonomic computing, multi-agent systems and frameworks in developing
applications for mobile devices, emphasizing autonomic management of web
services use. Aiming the consolidation of the investigated concepts and provide
a guide to facilitate the development of self-adapting applications, a framework
is proposed to help in applications development for mobile computing. The
development of two applications illustrates the framework appliance. The first
application is a system that presents the risk of mass sliding in an area
through a susceptibility calculation based on information of climate, topography
and soil. The second application is part of a travel agency system that defines
a travel package that best fit user needs.
[12_PhD_ridolfi]*
Lorenzo Francesco Giovanni Gino Maria RIDOLFI. Construção de espaços de cor
euclidianos e perceptualmente uniformes com base na fórmula CIEDE2000. [Title in
English: Construction of Perceptually Uniform Euclidean Color Spaces based on
the CIEDE2000 formula]. Ph.D. Thesis. Port. Presentation: 13/06/12. 93 p.
Advisor: Marcelo Gattass.
Abstract: In recent years, various color difference formulas were developed
for the CIELAB space, such as CMC, CIE94 and CIEDE2000. Although these formulas
have achieved greater accuracy in perceptual measurement between colors, many
applications cannot take advantage of this greater precision, because the
Euclidean distances in CIELAB are not isometric in accordance with these new
formulas. Thus, applications such as gamut mapping and color interpolation need
a color space that is isometric in relation to the latest color difference
formulas. This paper studies the mapping of the CIELAB space, particularly the
ab plane of this space according to the metrics of the CIEDE2000 formula,
through multidimensional scaling (MDS) techniques, more specifically ISOMAP and
an optimization based on Sammon Mapping.
[12_MSc_arruda]
Marcelo Medeiros ARRUDA.
Visualização volumétrica de dados sísmicos utilizando
um modelo de iluminação por oclusão direcional. [Title
in English: Visualization of seismic volumetric data using a directional
occlusion shading model].
M.Sc. Diss. Port. Presentation: 05/09/12. 66 p. Advisor: Waldemar Celes
Filho.
Abstract: The Interpretation of seismic volumetric data has a major
importance for the oil and gas industry. Sinse these data types have a
volumetric character mode, identify and select attributes present in this struct
become a difficult task. Furthermore, the high-frequency nois and depth
information typically found in this type of data, increases the complexity of
their manipulation and visualization. Due to these characteristics, the geometry
of 3D seismic data is very complex and is necessary more realistic light model
to perform the illumination of the seismic volume. This work consists of
performing a volumetric visualization of seismic data based on ray tracing
algorithm, using an illumination model by directional occlusion, computing the
ambiente light attenuated by the elements in the light trajetory for all
elements in the volume. Thus, we emphasize
the geometry of the seismic data, especially the depth cues and spatial
relationship. The proposed algorithm was full implemented on graphics card,
allowing manipulation at interactive rates, without any pre-processing.
[12_PhD_jacyntho]
Mark Douglas de Azevedo
JACYNTHO. Um modelo de bloqueio multigranular para RDF. [Title
in English: A Multigranularity locking model for RDF].
Ph.D. Thesis. Port. Presentation: 02/02/12. 277 p. Advisor: Daniel Schwabe.
Abstract: Client sessions in Web applications are organized as transactions
involving requests that read and write shared data. Executing concurrently,
these sessions may invalidate each other's data. The advent of Linked Data is
spurring the deployment of applications that use the RDF data model at the
information tier. In addition to querying RDF data, there is also the
requirement for online updates with suitable concurrency control. This work
presents a locking model, inspired by the multigranularity locking protocol, to
address isolation between transactions
that manipulate RDF data, considering four concurrency-related issues, namely:
lost updates, dirty reads, non-repeatable reads and phantom reads. Four
"hierarchically" related granules are offered, as well as new read and write
lock modes, specifically created for the RDF data model. Finally, the
performance assessment of the proposed locking model is done through simulation.
[12_MSc_viana]
Marx Leles VIANA. JSAN: um framework para simulação de agentes normativos. [Title
in English: JASAN: a framework for simulation of normative agents].
M.Sc. Diss. Port. Presentation: 03/04/12. 68 p. Advisor: Carlos José
Pereira da Lucena.
Abstract: Open multi-agent systems are societies in which autonomous and
heterogeneous agents can work towards similar or different ends. In order to
cope with the heterogeneity, autonomy and diversity of interests among the
different members, those systems establish a set of norms that is used as a
mechanism of social control to ensure a desirable social order in which agents
work together. Such norms regulate the behaviour of the agents by defining
obligations, permissions and prohibitions. Moreover, norms may give stimulus to
their fulfillment by defining rewards and may discourage their violation by
stating punishments. Although norms are promising mechanisms to regulate agents’
behavior, we should take into account that agents are autonomous entity, so they
must be free to decide to fulfill or violate each norm. In this way, agents can
use different strategies when deciding to achieve their goals and comply with
the norms addressed to themselves. On one hand, agents might choose to achieve
their goals without concerning with their norms, i.e., without concerting with
the rewards they could receive if they fulfill the norms and the punishments
they will receive for violating them. On the other hand, some agents will choose
to comply with all the norms although some of their goals may not be achieved.
In this context, this work proposes a framework for simulating normative agents
that provides the necessary mechanisms to understand the impacts of norms on
agents that adopt some of those strategies to deal with norms. The applicability
of the framework will be evaluated in two scenarios: the first in the context of
prevention of crimes and the second is related to the mission of rescuing
civilians who are at risk areas.
[12_MSc_soares]
Michele dos Santos SOARES.
Projetos de jogos educativos 2D de aventura usando
Lua. [Title
in English: Design of 2D educational games of adventures using Lua].
M.Sc. Diss. Port. Presentation: 04/09/12. 72 p. Advisor: Bruno Feijó.
Abstract: The computer can be considered an ally of the cognitive
development of students, especially by enabling the development of a work that
adapts itself to different learning rhythms. in the process of teaching
mathematics, for instance, educational software can contributed to create a
scenario that offers opportumities for the student to associate mathematical
concepts with his/her quotidian experiences, without losing funny aspects and
motivation. However, a problem faced by educators in the use of educational
software is limited flexibility of the avaliable systems. In most of the
educational games in the market, one can only experience static contents and
rigid structures. An educational game with static contents acts within a
specific context and once the user learns those contents the game fail to
providechallanges to the player and loses all the fun. As an attempt to
alleviate the consequences of those problems, this work presents a framwork for
the design of education 2D games in the adventure style that facilitates the
definition of objects and behaviors by educators
and artists, from the integration of a lua script layer with a 2D game engine.
[12_MSc_pena]
Rafael Antônio
Pinto PENA.
Suporte semântico à publicação de conteúdo jornalístico na Web. [Title
in English: Semantic support for the publication of news content on the Web].
M.Sc. Diss. Port. Presentation: 07/02/12. 105 p. Advisor: Daniel Schwabe.
Abstract: The use of journalistic content on the web increases every day.
However, much of the content is still produced according to the paradigms of
print media. Parallel to this, the Semantic Web or Web 3.0, adds a layer of "intelligence"
to the Web, where computers are able to extract meaning from the visited web
content, and consequently, process them. This work defined and tested a model of
publishing news content supported by the Semantic Web. It was developed as a
support tool for content producers, in order to improve the process of
construction of news stories for the web, using the suggestion of narrative
patterns and objective information that support the desired type of narrative. A
case study was conducted to evaluate the proposed model in a real environment. A
group of journalists used the developed tool, with unanimous acceptance.
[12_MSc_nasser]
Rafael Barbosa
NASSER.
McCloud Service Framework: arcabouço para desenvolvimento de serviços
baseados na simulação de Monte Carlo na Cloud. [Title
in English: McCloud Service Framework: development services of Monte Carlo
simulation in the Cloud].
M.Sc. Diss. Port. Presentation: 04/04/12. 107 p. Advisor: Karin Koogan
Breitman.
Abstract: The investment in computing infrastructure to attend
seasonal demand or processing peak can generate financial waste, because the
most of the time these resources are idle. In addition, in many solutions the
response time are critical to attend business requirements, which often, turn
the solution economically unviable. In this scenario it is essential intelligent
allocation of computing resources according to the demand for processing,
allocation and cost of business requirements. The Monte Carlo Simulation is a
statistical method widely used to solve a wide range of scientific and
engineering problems. When applied to real problems usually have the challenges
mentioned. Cloud Computing is an alternative to providing on-demand computing
resources, generating economies of scale unprecedented and almost infinite
scalability. Aligning a modern architecture to the cloud is possible to
encapsulate functionality and offer a range of services that would previously
have been restricted to specific areas. In this paper we are interested in
building a generic framework, that may provide a range of services based on
Monte Carlo, make rational use of the elasticity provided by the cloud in order
to achieve better levels of efficiency and reuse.
[12_MSc_waintraub]
Tatiana WAINTRAUB.
Modelagem da calçada de Copacabana. [Title in English: Modeling the
Copacabana sidewalk pavement]. M.Sc. Diss. Port. Presentation: 06/08/12. 42 p. Advisor: Waldemar Celes Filho.
Abstract: In
this dissertation, we propose an unsupervised method to model the Copacabana
beach sidewalk pavement, and Portuguese pavements in general. Given a black and
white source image, the proposed method outputs the geometry of all individual
stones that compose the
pavement. Different from previous mosaic techniques, we focus on capturing the
particularities of such pavements: stones (tiles) follow the edges, being mostly
represented by irregular quadrilaterals with no particular orientation. A set of
experiments demonstrates the effectiveness and stability of our proposal.
[12_Msc_toledo]
Thiago Marques TOLEDO.
Visualizando campos vetoriais em superfícies. [Title in English: Visualizing
vector fields over surfaces]. M.Sc. Diss. Port. Presentation: 15/05/12. 51 p. Advisor: Waldemar Celes Filho.
Abstract: Vector fields are common results
of physics simulators. Simulations over black-oil reservoirs, for instance, can
generate oil, water and gas flow data. For a better understanding of such data,
however, it’s interesting to use a visualization technique that allows a better
identification of local characteristics and global tendencies of the field. This
work proposes a technique for visualization of 3D vector fields that is
GPU-based and uses the 2D line integral convolution (LIC) algorithm to visualize
the component tangential to the surface projected on screen space. Data related
to magnitude and normal component are presented through a 2-dimensional color
scale. A simple scheme based on randomly generated texture coordinates is
proposed to fixate the resulting LIC image to the model, avoiding flickering
during model manipulation and eliminating the need for a solid 3D texture noise.
For animation, we adjust the use of filters to ensure that the animation speed
varies in accordance to the field magnitude. To enhance the final image, the LIC
algorithm is applied in two passes and the result is put through a high-pass
filter. The framework developed as part of this work has been applied in the
context of visualizing flow in black-oil reservoir models and height gradients
in terrains. In the specific case of reservoirs, a variation from the main
technique is proposed to allow simultaneous visualization of oil, gas and water
flows.
[12_MSc_leite]
Vanessa Rodrigues Coelho LEITE. Uma análise da classificação de litologias
utilizando SVM, MLP e métodos ensemble. [Title in English: An analysis of
lithology classification using SVM, MLP and ensemble methods]. M.Sc. Diss. Port.
Presentation: 13/07/12. 79 p. Advisor: Marcelo Gattass.
Abstract:
Lithology classification is an important task in oil reservoir characterization,
one of its major purposes is to support well planning and drilling activities.
Therefore, faster and more effective classification algorithms will increase the
speed and reliability of decisions made by geologists and geophysicists. This
work analises ensemble methods applied to automatic lithology classification.
For this, we performed a comparison between
single classifiers (Support Vector Machine and Multilayer Perceptron) and these
classifiers with ensemble methods (Bagging and Boost). Thus, we conclude with a
comparative evaluation of techniques and present the trade-off in using Ensemble
methods to replace single classifiers.
[12_MSc_fusco]
Victor Sá Freire FUSCO. Componentes de software com suporte a fluxo de dados.
[Title in English: Software components with support for data streams]. M.Sc.
Diss. Port. Presentation: 27/04/12. 83 p. Advisor: Renato Fontoura de Gusmão
Cerqueira.
Abstract:
Component-based software development is a topic that has attracted attention in
recent years. This technique allows the construction of complex software systems
in a quick and structured way. Several component models have been proposed by
the industry and the academy. The majority of these component models adopt
Remote Procedure Calls as their basic communication mechanism.The CORBA
Component Model is the only one from the surveyed models that has a work in
progress to support communication over data streams. This support proves to be
of great importance in systems that must deal with data from sensors and systems
that deal with audio and video transmission. The main goal of this work is to
propose an architecture that enables the middleware Software Component System (SCS)
to support applications that require data streaming. To this end, the SCS
component model was extended to support stream ports. As evaluation, this work
presents some experimental results of performance and scalability, as well as an
application that exercises the needs of the CSBase's algorithm ow executor, a
framework used to build systems for grid computing.
[12_Msc_fernandes]
William Paulo Ducca FERNANDES.
Quotation extraction for Portuguese. [Title in
Portuguese: Extração de citações para o Português]. M.Sc. Diss. Eng. Presentation: 09/04/12. 59 p. Advisor: Ruy
Luiz Milidiú.
Abstract: Quotation Extraction consists of identifying
quotations from a text and associating them to their authors. In this work, we
present a Quotation Extraction system for Portuguese. Quotation Extraction has
been previously approached using different techniques and for several languages.
Our proposal differs from previous work since we use Machine Learning to
automatically build specialized rules instead of human-derived rules. Machine
Learning models usually present stronger generalization power compared to
human-derived models. In addition, we are able to easily adapt our model to
other languages, needing only a list of verbs of speech for a given language.
The previously proposed systems would probably need a rule set adaptation to
correctly classify the quotations, which would be time consuming. We tackle the
Quotation Extraction task using one model for the Entropy Guided Transformation
Learning algorithm and another one for the Structured Perceptron algorithm. In
order to train and evaluate the system, we have build the GloboQuotes corpus,
with news extracted from the globo.com portal. We add part-of-speech tags to the
corpus using a state-of-the-art tagger. The Structured Perceptron based on
weighted interval scheduling obtains an Fß=1 score of 76.80%.