Consultas por palavras-chave permitem o acesso fácil a dados XML, uma vez que não exigem que o usuário aprenda uma linguagem de consulta estruturada nem estude possíveis esquemas de dados complexos. Com isso, vários motores de busca XML foram propostos para permitir a extração de fragmentos XML relevantes para consultas por palavras-chave. No entanto, esses motores de busca tratam as expressões temporais da mesma forma que qualquer outra palavra-chave. Essa abordagem ocasiona inúmeros problemas, como por exemplo, considerar como casamentos para uma expressão temporal nodos do domínio preço ou código. Este trabalho descreve TPI (Two Phase Interception), uma abordagem que permite o suporte a consultas temporais por palavras-chave em documentos XML orientados a dados. O suporte a consultas temporais é realizado através de uma camada adicional de software que executa duas interceptações no processamento de consultas, realizado por um motor de busca XML. Esta camada adicional de software é responsável pelo tratamento adequado das informações temporais presentes na consulta e no conteúdo dos documentos XML. O trabalho ainda especifica TKC (Temporal Keyword Classification), uma classificação de consultas temporais que serve de guia para qualquer mecanismo de consulta por palavras-chave, inclusive TPI. São apresentados os algoritmos de mapeamento das diferentes formas de predicados temporais por palavras-chave, especificadas em TKC, para expressões relacionais a fim de orientar a implementação do processamento das consultas temporais. É proposto um índice temporal e definidas estratégias para identificação de caminhos temporais, desambiguação de formatos de valores temporais, identificação de datas representadas por vários elementos e identificação de intervalos temporais. São demonstrados experimentos que comparam a qualidade, o tempo de processamento e a escalabilidade de um motor de busca XML com e sem a utilização de TPI. A principal contribuição desse trabalho é melhorar significativamente a qualidade dos resultados de consultas temporais por palavras-chave em documentos XML. / Keyword queries enable users to easily access XML data, since the user does not need to learn a structured query language or study possibly complex data schemas. Therewith, several XML search engines have been proposed to extract relevant XML fragments in response to keyword queries. However, these search engines treat the temporal expressions as any other keyword. This approach may lead to several problems. It could, for example, consider prices and codes as matches to a temporal expression. This work describes TPI (Two Phase Interception), an approach that supports temporal keyword queries on data-centric XML documents. The temporal query support is performed by adding an additional software layer that executes two interceptions in the query processing performed by a XML search engine. This additional software layer is responsible for the adequate treatment of the temporal expressions contained in the query and in the contents of the XML documents. This work also specifies TKC (Temporal Keyword Classification), a temporal query classification to be used as guidance for any keyword query mechanism, including TPI. We present the algorithms for mapping different temporal predicates expressed by keywords to relational expressions in order to guide the implementation of the temporal query processing. We propose a temporal index together with strategies to perform temporal path identification, format disambiguation, identification of dates represented by many elements and detection of temporal intervals. This work also reports on experiments which evaluate quality, processing time and scalability of an XML search engine with TPI and without TPI. The main contribution of this work is the significant improvement in the quality of the results of temporal keyword queries on XML documents.

Consultas por palavras-chave permitem o acesso fácil a dados XML, uma vez que não exigem que o usuário aprenda uma linguagem de consulta estruturada nem estude possíveis esquemas de dados complexos. Com isso, vários motores de busca XML foram propostos para permitir a extração de fragmentos XML relevantes para consultas por palavras-chave. No entanto, esses motores de busca tratam as expressões temporais da mesma forma que qualquer outra palavra-chave. Essa abordagem ocasiona inúmeros problemas, como por exemplo, considerar como casamentos para uma expressão temporal nodos do domínio preço ou código. Este trabalho descreve TPI (Two Phase Interception), uma abordagem que permite o suporte a consultas temporais por palavras-chave em documentos XML orientados a dados. O suporte a consultas temporais é realizado através de uma camada adicional de software que executa duas interceptações no processamento de consultas, realizado por um motor de busca XML. Esta camada adicional de software é responsável pelo tratamento adequado das informações temporais presentes na consulta e no conteúdo dos documentos XML. O trabalho ainda especifica TKC (Temporal Keyword Classification), uma classificação de consultas temporais que serve de guia para qualquer mecanismo de consulta por palavras-chave, inclusive TPI. São apresentados os algoritmos de mapeamento das diferentes formas de predicados temporais por palavras-chave, especificadas em TKC, para expressões relacionais a fim de orientar a implementação do processamento das consultas temporais. É proposto um índice temporal e definidas estratégias para identificação de caminhos temporais, desambiguação de formatos de valores temporais, identificação de datas representadas por vários elementos e identificação de intervalos temporais. São demonstrados experimentos que comparam a qualidade, o tempo de processamento e a escalabilidade de um motor de busca XML com e sem a utilização de TPI. A principal contribuição desse trabalho é melhorar significativamente a qualidade dos resultados de consultas temporais por palavras-chave em documentos XML. / Keyword queries enable users to easily access XML data, since the user does not need to learn a structured query language or study possibly complex data schemas. Therewith, several XML search engines have been proposed to extract relevant XML fragments in response to keyword queries. However, these search engines treat the temporal expressions as any other keyword. This approach may lead to several problems. It could, for example, consider prices and codes as matches to a temporal expression. This work describes TPI (Two Phase Interception), an approach that supports temporal keyword queries on data-centric XML documents. The temporal query support is performed by adding an additional software layer that executes two interceptions in the query processing performed by a XML search engine. This additional software layer is responsible for the adequate treatment of the temporal expressions contained in the query and in the contents of the XML documents. This work also specifies TKC (Temporal Keyword Classification), a temporal query classification to be used as guidance for any keyword query mechanism, including TPI. We present the algorithms for mapping different temporal predicates expressed by keywords to relational expressions in order to guide the implementation of the temporal query processing. We propose a temporal index together with strategies to perform temporal path identification, format disambiguation, identification of dates represented by many elements and detection of temporal intervals. This work also reports on experiments which evaluate quality, processing time and scalability of an XML search engine with TPI and without TPI. The main contribution of this work is the significant improvement in the quality of the results of temporal keyword queries on XML documents.

abstract: Speech recognition and keyword detection are becoming increasingly popular applications for mobile systems. While deep neural network (DNN) implementation of these systems have very good performance, they have large memory and compute resource requirements, making their implementation on a mobile device quite challenging. In this thesis, techniques to reduce the memory and computation cost of keyword detection and speech recognition networks (or DNNs) are presented. The first technique is based on representing all weights and biases by a small number of bits and mapping all nodal computations into fixed-point ones with minimal degradation in the accuracy. Experiments conducted on the Resource Management (RM) database show that for the keyword detection neural network, representing the weights by 5 bits results in a 6 fold reduction in memory compared to a floating point implementation with very little loss in performance. Similarly, for the speech recognition neural network, representing the weights by 6 bits results in a 5 fold reduction in memory while maintaining an error rate similar to a floating point implementation. Additional reduction in memory is achieved by a technique called weight pruning, where the weights are classified as sensitive and insensitive and the sensitive weights are represented with higher precision. A combination of these two techniques helps reduce the memory footprint by 81 - 84% for speech recognition and keyword detection networks respectively. Further reduction in memory size is achieved by judiciously dropping connections for large blocks of weights. The corresponding technique, termed coarse-grain sparsification, introduces hardware-aware sparsity during DNN training, which leads to efficient weight memory compression and significant reduction in the number of computations during classification without loss of accuracy. Keyword detection and speech recognition DNNs trained with 75% of the weights dropped and classified with 5-6 bit weight precision effectively reduced the weight memory requirement by ~95% compared to a fully-connected network with double precision, while showing similar performance in keyword detection accuracy and word error rate. / Dissertation/Thesis / Masters Thesis Computer Science 2016

The purpose of this thesis is to investigate organizational quality assurance through the international testing standard ISO 29119. The focus will be on how an organization carries out testing processes and designs and implements test cases. Keyword driven testing is a test composition concept in ISO 29119 and suitable for automation. This thesis will answer how keyword driven testing can facilitate the development of maintainable test cases and support test automation in an agile organization. The methodology used was a qualitative case study including semi-structured interviews and focus groups with agile business units within Scania IT. Among the interview participants were developers, test engineers, scrum masters and a unit manager. The results describe testing practices carried out in several agile business units, maintainability issues with test automation and general ideas of how test automation should be approached. Common issues with test automation were test cases failing due to changed test inputs, inexperience with test automation frameworks and lack of resources due to project release cycle. This thesis concludes that keyword driven testing has the potential of solving several maintainability issues with test cases breaking. However, the practicality and effectiveness of said potential remain unanswered. Moreover, successfully developing an automated keyword driven testing framework requires integration with existing test automation tools and considering the agile organizational circumstances.

Consultas por palavras-chave permitem o acesso fácil a dados XML, uma vez que não exigem que o usuário aprenda uma linguagem de consulta estruturada nem estude possíveis esquemas de dados complexos. Com isso, vários motores de busca XML foram propostos para permitir a extração de fragmentos XML relevantes para consultas por palavras-chave. No entanto, esses motores de busca tratam as expressões temporais da mesma forma que qualquer outra palavra-chave. Essa abordagem ocasiona inúmeros problemas, como por exemplo, considerar como casamentos para uma expressão temporal nodos do domínio preço ou código. Este trabalho descreve TPI (Two Phase Interception), uma abordagem que permite o suporte a consultas temporais por palavras-chave em documentos XML orientados a dados. O suporte a consultas temporais é realizado através de uma camada adicional de software que executa duas interceptações no processamento de consultas, realizado por um motor de busca XML. Esta camada adicional de software é responsável pelo tratamento adequado das informações temporais presentes na consulta e no conteúdo dos documentos XML. O trabalho ainda especifica TKC (Temporal Keyword Classification), uma classificação de consultas temporais que serve de guia para qualquer mecanismo de consulta por palavras-chave, inclusive TPI. São apresentados os algoritmos de mapeamento das diferentes formas de predicados temporais por palavras-chave, especificadas em TKC, para expressões relacionais a fim de orientar a implementação do processamento das consultas temporais. É proposto um índice temporal e definidas estratégias para identificação de caminhos temporais, desambiguação de formatos de valores temporais, identificação de datas representadas por vários elementos e identificação de intervalos temporais. São demonstrados experimentos que comparam a qualidade, o tempo de processamento e a escalabilidade de um motor de busca XML com e sem a utilização de TPI. A principal contribuição desse trabalho é melhorar significativamente a qualidade dos resultados de consultas temporais por palavras-chave em documentos XML. / Keyword queries enable users to easily access XML data, since the user does not need to learn a structured query language or study possibly complex data schemas. Therewith, several XML search engines have been proposed to extract relevant XML fragments in response to keyword queries. However, these search engines treat the temporal expressions as any other keyword. This approach may lead to several problems. It could, for example, consider prices and codes as matches to a temporal expression. This work describes TPI (Two Phase Interception), an approach that supports temporal keyword queries on data-centric XML documents. The temporal query support is performed by adding an additional software layer that executes two interceptions in the query processing performed by a XML search engine. This additional software layer is responsible for the adequate treatment of the temporal expressions contained in the query and in the contents of the XML documents. This work also specifies TKC (Temporal Keyword Classification), a temporal query classification to be used as guidance for any keyword query mechanism, including TPI. We present the algorithms for mapping different temporal predicates expressed by keywords to relational expressions in order to guide the implementation of the temporal query processing. We propose a temporal index together with strategies to perform temporal path identification, format disambiguation, identification of dates represented by many elements and detection of temporal intervals. This work also reports on experiments which evaluate quality, processing time and scalability of an XML search engine with TPI and without TPI. The main contribution of this work is the significant improvement in the quality of the results of temporal keyword queries on XML documents.

Submitted by Divisão de Documentação/BC Biblioteca Central (ddbc@ufam.edu.br) on 2017-08-22T19:40:10Z No. of bitstreams: 2 Tese - Péricles Silva de Oliveira.pdf: 1875380 bytes, checksum: 014ba89b7fe1929a1461c9d8d3959416 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Divisão de Documentação/BC Biblioteca Central (ddbc@ufam.edu.br) on 2017-08-22T19:40:26Z (GMT) No. of bitstreams: 2 Tese - Péricles Silva de Oliveira.pdf: 1875380 bytes, checksum: 014ba89b7fe1929a1461c9d8d3959416 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Divisão de Documentação/BC Biblioteca Central (ddbc@ufam.edu.br) on 2017-08-22T19:40:44Z (GMT) No. of bitstreams: 2 Tese - Péricles Silva de Oliveira.pdf: 1875380 bytes, checksum: 014ba89b7fe1929a1461c9d8d3959416 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-08-22T19:40:44Z (GMT). No. of bitstreams: 2 Tese - Péricles Silva de Oliveira.pdf: 1875380 bytes, checksum: 014ba89b7fe1929a1461c9d8d3959416 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-04-28 / Several systems proposed for processing keyword queries over relational databases rely on the generation and evaluation of Candidate Networks (CNs), i.e., networks of joined database relations that, when processed as SQL queries, provide a relevant answer to the input keyword query. Although the evaluation of CNs has been extensively addressed in the literature, problems related to efficiently generating meaningful CNs have received much less attention. To generate useful CNs is necessary to automatically locating, given a handful of keywords, relations in the database that may contain relevant pieces of information, and determining suitable ways of joining these relations to satisfy the implicit information need expressed by a user when formulating her query. In this thesis, we present two main contributions related to the processing of Candidate Networks. As our first contribution, we present a novel approach for generating CNs, in which possible matchings of the query in database are efficiently enumerated at first. These query matches are then used to guide the CN generation process, avoiding the exhaustive search procedure used by current state-of-art approaches. We show that our approach allows the generation of a compact set of CNs that leads to superior quality answers, and that demands less resources in terms of processing time and memory. As our second contribution, we initially argue that the number of possible Candidate Networks that can be generated by any algorithm is usually very high, but that, in fact, only very few of them produce answers relevant to the user and are indeed worth processing. Thus, there is no point in wasting resources processing useless CNs. Then, based on such an argument, we present an algorithm for ranking CNs, based on their probability of producing relevant answers to the user. This relevance is estimated based on the current state of the underlying database using a probabilistic Bayesian model we have developed. By doing so we are able do discard a large number of CNs, ultimately leading to better results in terms of quality and performance. Our claims and proposals are supported by a comprehensive set of experiments we carried out using several query sets and datasets used in previous related work and whose results we report and analyse here. / Sem resumo.

Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / Existe um perfil de torcedor de futebol que tem como prÃtica torcer por um time de seu estado e por outro, sediado em uma regiÃo diferente. Ele à pejorativamente chamado de âtorcedor mistoâ. Essa alcunha foi criada por torcedores que nÃo concordavam com esse modo de se torcer. O perfil de torcedor âmistoâ existe em praticamente todos os estados do Brasil. Essa realidade fez com que, desde 2008, alguns grupos de torcedores, contrÃrios Ãquele modo de se torcer, organizassem faixas, camisetas, vÃdeos e comunidades virtuais para mostrar seu ponto de vista e estimular os âmistosâ a torcerem somente por um time de seu estado. Esse movimento, conhecido como âanti-mistosâ, teve inÃcio na regiÃo nordestina, tendo como uma de suas sentenÃas de maior repercussÃo aquela que classifica os âmistosâ como âA vergonha do Nordesteâ. Este trabalho tem como objetivo compreender algumas das situaÃÃes sociais, histÃricas e subjetivas que estimulam o advento deste perfil de torcedor, tomando como recorte a RegiÃo Nordeste. Analisando o futebol brasileiro como um campo (Bourdieu), constatou-se que os times que detÃm os principais tipos de capital e sÃo mais ativos nas instituiÃÃes desse campo - sendo dominantes - sÃo clubes de Rio de Janeiro e SÃo Paulo. Os clubes nordestinos e das demais regiÃes ocupam a condiÃÃo de dominados. Essas situaÃÃes tornam os clubes dominantes bastante atraentes ao fà do esporte, que tambÃm và na escolha do time uma forma de conseguir reconhecimento e distinÃÃo. As principais emissoras de TV do paÃs sÃo sediadas nos estados de Rio de Janeiro e SÃo Paulo. AtravÃs de sua rede de afiliadas pelo territÃrio nacional, diariamente levavam aos fÃs de futebol por todo paÃs considerÃvel quantidade de informaÃÃes sobre os clubes daqueles estados, incluindo transmissÃes ao vivo de suas participaÃÃes em torneios brasileiros e internacionais. Essa presenÃa constante na mÃdia ajuda esses clubes a se tornarem simbolicamente muito prÃximos do dia a dia do torcedor nordestino, mesmo que geograficamente distantes. Essa proximidade contribui para que esse torcedor nÃo interprete aqueles clubes como algo que âpertence aos outrosâ (argumento recorrente entre os âanti-mistosâ), mas sim alguma coisa da qual ele tambÃm pode fazer parte. / There is a profile of soccer fans which usually supports a team from its state and another club, based in a different region. This kind of fan is pejoratively called âmistoâ (mixed). This appellation was created by fans who disagree with this way of cheering. The fan called âmistoâ exists in all the Brazilian states. Because of this, since 2008 some groups of fans, non âmistosâ, started to organize sashes, shirts, videos and virtual communities to show their point of view and encourage the âmistosâ to support only a team from its state. This mobilization, known âanti-mistoâ, started in Northeast. One of its most popular watchwords is âA vergonha do Nordesteâ (The shame of Northeast). This report aims to understand some of the social, historic and subjective situations that stimulated the advent of the fans called âmistoâ, analysing the Northeast case. Analysing Brazilian soccer as a field (Bourdieu), it could find some teams that holds the main types of capital and are more actives in the institutions. They are dominants. These teams belong to Rio de Janeiro and SÃo Paulo. The clubs from Northeast and other regions are dominated. The dominants are very attractive for the fans, who also found in the choice for a soccer team a way to conquest recognition and distinction. The main TV networks in Brazil are based in SÃo Paulo and Rio de Janeiro. Using their net of affiliated TVs across the country, they give daily to soccer fans from every regions a lot of information about the teams from SÃo Paulo and Rio de Janeiro, including live transmit from their participation in national and international competitions. This constant presence in media makes these teams symbolically close to the routine of the Northeast fan, although geographically distant. Because of this proximity, the âmistoâ doesnât see that teams as clubs which belong to other people (an âanti-mistoâ recurrent argument), but as something to which he could belong.

This study aims to test Robertson’s lexical transfer principle, which posits that Chinese learners use demonstratives (particularly this) and the numeral one as markers of definiteness and indefiniteness. This is tested by analysing Chinese learners’ written production collected from the Spoken and Written English Corpus of Chinese Learners 2.0 (SWECCL 2.0). The purpose is to understand the variation of article usage by adult Chinese learners of English. More specifically, the study examines to what extent articles, possessive and demonstrative pronouns are used in Chinese learners’ English and how definite and indefinite articles are used by the Chinese learners. Findings suggest that Robertson’s lexical transfer principle was corroborated by the present study. In addition, Chinese learners prefer to use demonstrative determiners, the possessive determiner our, and the numeral one to perform the function of marking definiteness and indefiniteness. In particular, the learners try to use the demonstrative determiners that and this in the anaphoric function instead of the definite article, and the demonstrative determiner those is frequently used in the cataphoric function. What is more, the learners use the numeral one as a marker of indefiniteness, and it is also used as a marker of definiteness in the anaphoric function. Further, the possessive determiner our is used as a marker of definiteness in larger situation uses referring to something unique. To this end, the study is able to show that the definite article is used to perform the function of marking indefiniteness, and in some particular contexts the definite article functions as a Chinese specifier in Chinese learners’ English. Also, the indefinite article is frequently used in quantifier phrases but is rarely used in other functions. There are three main reasons that may explain why Chinese learners use determiners variously. Firstly, the choice of determiners by Chinese learners is influenced by linguistic contexts. Secondly, because of learning strategies, Chinese learners try to ignore the anaphoric function and cataphoric function that they are not yet ready to process in article usage. Thirdly, interlanguage grammar influences the optionality in the use of articles.

Cette thèse s’intègre dans le cadre du projet européen EUREKA "High Dynamic Range - Low NoiseCMOS imagers", qui a pour but de développer de nouvelles approches de fabrication de capteursd’images CMOS à haute performance. L’objectif de la thèse est la conception d’un système de visiontemps réel à grande gamme dynamique (HDR). L’axe principal sera la reconstruction, en temps réelet à la cadence du capteur (60 images/sec), d’une vidéo à grande dynamique sur une architecturede calcul embarquée.La plupart des capteurs actuels produisent une image numérique qui n’est pas capable de reproduireles vraies échelles d’intensités lumineuses du monde réel. De la même manière, les écrans, impri-mantes et afficheurs courants ne permettent pas la restitution effective d’une gamme tonale étendue.L’approche envisagée dans cette thèse est la capture multiple d’images acquises avec des tempsd’exposition différents permettant de palier les limites des dispositifs actuels.Afin de concevoir un système capable de s’adapter temporellement aux conditions lumineuses,l’étude d’algorithmes dédiés à la grande dynamique, tels que les techniques d’auto exposition, dereproduction de tons, en passant par la génération de cartes de radiances est réalisée. Le nouveausystème matériel de type "smart caméra" est capable de capturer, générer et restituer du contenu àgrande dynamique dans un contexte de parallélisation et de traitement des flux vidéos en temps réel / This thesis is a part of the EUREKA European project called "High Dynamic Range - Low NoiseCMOS imagers", which developped new approaches to design high performance CMOS sensors.The purpose of this thesis is to design a real-time high dynamic range (HDR) vision system. Themain focus will be the real-time video reconstruction at 60 frames/sec in an embedded architecture.Most of the sensors produce a digital image that is not able to reproduce the real world light inten-sities. Similarly, monitors, printers and current displays do not recover of a wide tonal range. Theapproach proposed in this thesis is multiple acquisitions, taken with different exposure times, to over-come the limitations of the standard devices.To temporally adapt the light conditions, the study of algorithms dedicated to the high dynamic rangetechniques is performed. Our new smart camera system is able to capture, generate and showcontent in a highly parallelizable context for a real time processing

La mise à l’échelle des technologies CMOS s’accompagne d’une réduction des tensions d’alimentation qui dégrade fortement la fonctionnalité des circuits RF. L’effet de substrat, conventionnellement considéré comme un effet parasite du transistor MOS, est ici exploité pour proposer des topologies de circuits capables de supporter un fonctionnement sous faible tension d’alimentation. Cette thématique est l’objet principale de ma thèse que j’ai débuté en Septembre 2007, supportée par une bourse MENRT. La pré-polarisation a permis de réduire de 1.2 V à 0.5 V la tension d’alimentation d’une structure cascode en technologie CMOS 0.13 µm. Une méthodologie de conception sous contrainte de faible consommation en puissance a été ensuite validée par les mesures. Il a été également démontré que la linéarité de circuits RF peut être optimisée par l’application d’une tension appropriée sur le substrat du transistor principal MOS. Le démonstrateur, un LNA, utilise un DAC pour l’ajustement de cette tension. Il accède ainsi au concept des circuits à contrôle numérique ou « digitally enhanced ». / Abstract