Big amounts of test data generated by the current used software testing tools (QA-C/QA-C++ and Cantata) contain a variety of different values. The variances cause enormous challenges in data aggregation and interpretation that directly affect generation of test metrics. Due to the circumstance of data processing, this master thesis introduces a database-centric test management framework for test metrics aims at centrally handling the big data as well as facilitating the generation of test metrics. Each test result will be individually parsed to be a particular format before being stored in a centralized database. A friendly front-end user interface is connected and synchronized with the database that allows authorized users to interact with the stored data. With a granularity tracking mechanism, any stored data will be systematically located and programmatically interpreted by a test metrics generator to create various kinds of high-quality test metrics. The automatization of the framework is driven by Jenkins CI to automatically and periodically performing the sequential operations. The technology greatly and effectively optimizes and reduces effort in the development, as well as enhance the performance of the software testing processes. In this research, the framework is only started at managing the testing processes on software-unit level. However, because of the independence of the database from levels of software testing, it could also be expanded to support software development at any level.

Intra-industry trade occurs when goods from the same industry category are both exported and imported. Types of intra-industry trade are identified, and theoretical models of intraindustry trade under conditions of imperfect competition are examined. The results of thirtyseven empirical studies on the determinants of intra-industry trade are analysed. Methods of measuring intra-industry trade and marginal intra-industry trade are discussed, and various measurement problems are dealt with. The extent of intra-industry trade in South Africa in 1992 and 1997 is measured, using the Grubel-Lloyd and Michaely indices. The BrUlhart indices are applied to measure marginal intra-industry trade. South Africa has a relatively low and stable level of intra-industry trade in manufactured goods: the GrubelLloyd index for 1997 is calculated to be 37 per cent. / Economics and Management Sciences / M.Com. (Economics)

Made available in DSpace on 2016-06-02T19:06:10Z (GMT). No. of bitstreams: 1 5631.pdf: 1787133 bytes, checksum: c363525148fa6a5fe71608e7a8ffcf4c (MD5) Previous issue date: 2012-09-28 / Universidade Federal de Sao Carlos / The Wireless Sensor Networks (WSNs) and Mobile Wireless Sensor Networks (MWSNs) are being increasingly used by different applications, such as monitoring of animals, monitoring of vital signs, environmental monitoring, surveillance and protection of critical infrastructure, leaking gas, among many others. Some of these applications are already making use of mobile sensor nodes, such as underwater monitoring, precision agriculture, among many others. Due to restricted resources of sensor nodes, especially in relation to energy consumption, the development for solutions based on WSN and MWSN becomes limited. The use of mobile sensor nodes, which typically has more computational resources, power and communication, can help to reduce the energy consumption of fixed nodes, increasing the lifetime of the network. Networks that use mobile sensor nodes (fixed and mobile) with different types of hardware are called Wireless Sensor Networks Heterogeneous Mobile. This paper presents the RAHMoN (Routing Algorithm for Heterogeneous Mobile Networks), which makes use of data aggregation technique to reduce the traffic transmissions on the network, hierarchy of nodes (clustering), and use of sensor nodes (fixed and mobile) that collaborate to deliver data to a sink node at high speed. In RAHMoN, the network is configured using the techniques of inundation (flooding) and inundation reverse (reverse flooding) to collect the fixed position of sensor nodes and form an adjacency matrix. This matrix helps to build routes for data delivery to the sink and is stored in the mobile sensor nodes. Results show that our solution can guarantee a high packages delivery rate, low latency and reduce the delay of packet delivery. The solution was compared with the WHISPER, present in the literature and also focused on the delivery of data to sink node at high speed. / As Redes de sensores sem Fio (RSSFs) e Redes de Sensores Sem Fio Móveis (RSSFMs) estão sendo cada vez mais utilizadas por diferentes aplicações, tais como: monitoramento de animais, monitoramento de sinais vitais, monitoramento ambiental, vigilância e proteção de infraestruturas críticas, vazamento de gás, dentre inúmeras outras. Algumas dessas aplicações já fazem uso de nós sensores móveis. Devido aos recursos restritos dos nós sensores, principalmente em relação ao consumo energético, o desenvolvimento de soluções baseadas em RSSF e RSSFM torna-se limitado. O uso de nós sensores móveis, que tipicamente têm mais recursos computacionais, de energia e de comunicação, pode ajudar a reduzir o consumo de energia dos nós fixos, aumentando o tempo de vida da rede. Redes que utilizam nós sensores (fixos e móveis) com diferentes tipos de hardware são denominadas Redes de Sensores Sem Fio Móvel Heterogênea. Neste trabalho é apresentado o RAHMoN (Routing Algorithm for Heterogeneous Mobile Networks), que faz uso da técnica de agregação de dados para reduzir o tráfego de transmissões na rede, da hierarquização de nós (clustering), da utilização de nós sensores (fixos e móveis) e de um sink em alta velocidade. No RAHMoN, a rede é configurada utilizando flooding e flooding reverse para coletar a posição dos nós sensores fixos e formar uma matriz de adjacência. Essa matriz auxilia na construção de rotas durante a entrega dos dados para o sink e será armazenada nos nós sensores móveis. Resultados de avaliação mostram que a nossa solução consegue garantir uma alta taxa de entrega de pacotes, diminuir a latência e reduzir o atraso de entrega dos pacotes. A solução foi comparada com o WHISPER, presente na literatura e também voltado à entrega de dados para o nó sink em alta velocidade.

Depuis plusieurs années, l’agrégation de données sont considérés comme un domaine émergent et prometteur tant dans le milieu universitaire que dans l’industrie. L’énergie et la capacité du réseau seront donc économisées car il y aura moins de transmissions de données. Le travail de cette thèse s’intéresse principalement aux fonctions d’agrégation Nous faisons quatre contributions principales. Tout d’abord, nous proposons deux nouvelles métriques pour évaluer les performances des fonctions d’agrégations vue au niveau réseau : le taux d’agrégation et le facteur d’accroissement de la taille des paquets. Le taux d’agrégation est utilisé pour mesurer le gain de paquets non transmis grâce à l’agrégation tandis que le facteur d’accroissement de la taille des paquets permet d’évaluer la variation de la taille des paquets en fonction des politiques d’agrégation. Ces métriques permettent de quantifier l’apport de l’agrégation dans l’économie d’énergie et de la capacité utilisée en fonction du protocole de routage considéré et de la couche MAC retenue. Deuxièmement, pour réduire l’impact des données brutes collectées par les capteurs, nous proposons une méthode d’agrégation de données indépendante de la mesure physique et basée sur les tendances d’évolution des données. Nous montrons que cette méthode permet de faire une agrégation spatiale efficace tout en améliorant la fidélité des données agrégées. En troisième lieu, et parce que dans la plupart des travaux de la littérature, une hypothèse sur le comportement de l’application et/ou la topologie du réseau est toujours sous-entendue, nous proposons une nouvelle fonction d’agrégation agnostique de l’application et des données devant être collectées. Cette fonction est capable de s’adapter aux données mesurées et à leurs évolutions dynamiques. Enfin, nous nous intéressons aux outils pour proposer une classification des fonctions d’agrégation. Autrement dit, considérant une application donnée et une précision cible, comment choisir les meilleures fonctions d’agrégations en termes de performances. Les métriques, que nous avons proposé, sont utilisées pour mesurer la performance de la fonction, et un processus de décision markovien est utilisé pour les mesurer. Comment caractériser un ensemble de données est également discuté. Une classification est proposée dans un cadre précis. / Wireless Sensor Networks (WSNs) have been regarded as an emerging and promising field in both academia and industry. Currently, such networks are deployed due to their unique properties, such as self-organization and ease of deployment. However, there are still some technical challenges needed to be addressed, such as energy and network capacity constraints. Data aggregation, as a fundamental solution, processes information at sensor level as a useful digest, and only transmits the digest to the sink. The energy and capacity consumptions are reduced due to less data packets transmission. As a key category of data aggregation, aggregation function, solving how to aggregate information at sensor level, is investigated in this thesis. We make four main contributions: firstly, we propose two new networking-oriented metrics to evaluate the performance of aggregation function: aggregation ratio and packet size coefficient. Aggregation ratio is used to measure the energy saving by data aggregation, and packet size coefficient allows to evaluate the network capacity change due to data aggregation. Using these metrics, we confirm that data aggregation saves energy and capacity whatever the routing or MAC protocol is used. Secondly, to reduce the impact of sensitive raw data, we propose a data-independent aggregation method which benefits from similar data evolution and achieves better recovered fidelity. Thirdly, a property-independent aggregation function is proposed to adapt the dynamic data variations. Comparing to other functions, our proposal can fit the latest raw data better and achieve real adaptability without assumption about the application and the network topology. Finally, considering a given application, a target accuracy, we classify the forecasting aggregation functions by their performances. The networking-oriented metrics are used to measure the function performance, and a Markov Decision Process is used to compute them. Dataset characterization and classification framework are also presented to guide researcher and engineer to select an appropriate functions under specific requirements.

Computers network in the hospital environments are central topics of discussion on the use of systems applied to health care to ensure data capture of vital signs. This study aims at analyzing the data capture model of the proposed sensor node, using the Modbus protocol communication standard in the acquisition of multi-parametric information of biological signs of patients. The research was conducted with experimental purpose to characterize the ModBus protocol in the RS485 serial network. It was also performed a systematic review to support the choice of the data model and serial communication standard in wired networks. Thus, a sensor node prototype (PIC18F26K20) was built to capture body temperature and heart rate in the wired communication network in which the protocols CAN and ModBUS were tested. The data from the sensor nodes were subjected to capture tests and sending the data to the central node, and displayed on portable platform (Smartphones). It was also observed the bandwidth characteristics and quality of the obtained data. The systematic review showed a trend in the use of the CAN protocol as wired communication standard for HealthCare activities, the application used in the experiment presented limitations. However, after the experiment using the ModBus protocol, this was adequate and easy to implement applications in the hospital environment, having a low-cost platform as a solution to that area. / Redes de computadores em ambientes hospitalares são temas centrais de discussão sobre a utilização de sistemas aplicados à atividade em saúde que garantam a captura dos dados de sinais vitais. O presente trabalho teve como objetivo analisar o modelo de captura dos dados do nodo sensor proposto, utilizando o padrão de comunicação do protocolo ModBUS na aquisição de informações multiparamétricas de sinais biológicos de pacientes. A pesquisa foi desenvolvida com propósito experimental para a caracterização do protocolo ModBUS sob a rede serial RS485. Foi realizada, ainda, uma revisão sistemática para subsidiar a escolha do modelo de dados e do padrão de comunicação serial em redes cabeadas. Para tanto, foi construído um protótipo de nodo sensor (PIC18F26K20) para captura de temperatura corporal e frequência cardíaca na comunicação da rede com fio na qual os protocolos CANbus e ModBUS foram testados. Os dados dos nodos sensores foram submetidos a testes de captura e envio dos dados ao nodo central, e exibidos em plataforma portáteis (Smartphones). Foram, ainda, observadas as características de largura de banda e qualidade dos dados obtidos. A revisão sistemática demonstrou uma tendência na utilização do protocolo CANbus como padrão de comunicação cabeado para as atividades HealthCare, a aplicação utilizada no experimento apresentou limitações. No entanto, após experimento utilizando o protocolo ModBus, este mostrou-se adequado e de fácil implementação em aplicações na área hospitalar, tendo uma plataforma de baixo custo como solução para a referida área.

Les recherches présentées dans ce mémoire s’inscrivent dans le cadre des réseaux decapteurs périodiques. Elles portent sur l’étude et la mise en oeuvre d’algorithmes et de protocolesdistribués dédiés à la gestion de données volumineuses, en particulier : la collecte, l’agrégation etla fouille de données. L’approche de la collecte de données permet à chaque noeud d’adapter sontaux d’échantillonnage à l’évolution dynamique de l’environnement. Par ce modèle le suréchantillonnageest réduit et par conséquent la quantité d’énergie consommée. Elle est basée surl’étude de la dépendance de la variance de mesures captées pendant une même période voirpendant plusieurs périodes différentes. Ensuite, pour sauvegarder plus de l’énergie, un modèled’adpatation de vitesse de collecte de données est étudié. Ce modèle est basé sur les courbes debézier en tenant compte des exigences des applications. Dans un second lieu, nous étudions unetechnique pour la réduction de la taille de données massive qui est l’agrégation de données. Lebut est d’identifier tous les noeuds voisins qui génèrent des séries de données similaires. Cetteméthode est basée sur les fonctions de similarité entre les ensembles de mesures et un modèle defiltrage par fréquence. La troisième partie est consacrée à la fouille de données. Nous proposonsune adaptation de l’approche k-means clustering pour classifier les données en clusters similaires,d’une manière à l’appliquer juste sur les préfixes des séries de mesures au lieu de l’appliquer auxséries complètes. Enfin, toutes les approches proposées ont fait l’objet d’études de performancesapprofondies au travers de simulation (OMNeT++) et comparées aux approches existantes dans lalittérature. / This thesis proposes novel big data management techniques for periodic sensor networksembracing the limitations imposed by wsn and the nature of sensor data. First, we proposed anadaptive sampling approach for periodic data collection allowing each sensor node to adapt itssampling rates to the physical changing dynamics. It is based on the dependence of conditionalvariance of measurements over time. Then, we propose a multiple level activity model that usesbehavioral functions modeled by modified Bezier curves to define application classes and allowfor sampling adaptive rate. Moving forward, we shift gears to address the periodic dataaggregation on the level of sensor node data. For this purpose, we introduced two tree-based bilevelperiodic data aggregation techniques for periodic sensor networks. The first one look on aperiodic basis at each data measured at the first tier then, clean it periodically while conservingthe number of occurrences of each measure captured. Secondly, data aggregation is performedbetween groups of nodes on the level of the aggregator while preserving the quality of theinformation. We proposed a new data aggregation approach aiming to identify near duplicatenodes that generate similar sets of collected data in periodic applications. We suggested the prefixfiltering approach to optimize the computation of similarity values and we defined a new filteringtechnique based on the quality of information to overcome the data latency challenge. Last butnot least, we propose a new data mining method depending on the existing K-means clusteringalgorithm to mine the aggregated data and overcome the high computational cost. We developeda new multilevel optimized version of « k-means » based on prefix filtering technique. At the end,all the proposed approaches for data management in periodic sensor networks are validatedthrough simulation results based on real data generated by periodic wireless sensor network.

Les médias sociaux occupent un rôle grandissant dans de nombreux domaines de notre vie quotidienne. Parmi d'autres, les réseaux sociaux tels que Facebook, Twitter, LinkedIn et Google+ dont la popularité a explosé ces dernières années, attirent des millions d'utilisateurs qui se communiquent, publient et partagent des informations et contenus à un rythme sans précédent. Outre les avantages reconnus, les réseaux sociaux ont également soulevé des problèmes divers. Nous sommes particulièrement intéressés par deux problèmes spécifiques : surcharge d'information et cloisonnement de données. Ces deux problèmes empêchent les utilisateurs d'exploiter pleinement et efficacement la richesse des informations poussées sur les réseaux sociaux. Les utilisateurs ont des difficultés pour filtrer tous les contenus reus, pour découvrir de nouveaux contenus au-delà de leurs réseaux personnels, et surtout pour partager les contenus intéressants avec leurs différents groupes d'intérêt. Pour aider les utilisateurs à surmonter ces difficultés, nous proposons une Approche centrée sur utilisateur et basée groupe pour filtrer et partager des données sociales. Cette nouvelle approche a un double objectif : (1) permettre aux utilisateurs d'agréger leurs données sociales en provenance de différents réseaux sociaux, d'en extraire des contenus de leur intérêt et (2) organiser et partager les contenus au sein de différents groupes. Les membres d'un groupe sont en outre en mesure de choisir quelle partie de leurs données à partager avec le groupe et définir collectivement les sujets d’intérêt de ce dernier. Pour implémenter l'approche proposée, nous spécifions une architecture de système comprenant plusieurs modules extensibles, et nous développons un prototype fonctionnel basé Web, appelé SoCoSys. Les résultats expérimentaux, obtenus des deux tests différents, valident les valeurs ajoutées de notre approche. / The social media have played an increasingly important role in many areas of our every day life. Among others, social network sites such as Facebook, LinkedIn, Twitter and Google+ have recently exploded in popularity by attracting millions of users, who communicate with each other, share and publish information and contents at an unprecedented rate. Besides the recognized advantages, social network sites have also raised various issues and challenges. We are particularly interested in two of them, information overload and "walled gardens". These two problems prevent the users from fully and efficiently exploiting thewealth of information available on social network sites. The users have difficulties to filter all incoming contents, to discover additional contents from outside of their friend circles, and importantly to share interesting contents with their different groups of interest. For helping the users to overcome such difficulties, we propose a User-centered and group- based approach for social data filtering and sharing. This novel approach has a twofold purpose : (1) allow the users to aggregate their social data from different social network sites, and to extract from those data the contents of their interest, and (2) organize and share the contents within different groups. The members of a group are moreover able to choose which part of their social data to share with the group, and collectively define its topics of interest. To achieve the proposed approach, we define a modular system architecture including a number of extensible modules, and accordingly build a working Web-based prototype, called SoCoSys. The experimental results, obtained from the two different tests, confirm the added values of our approach.

Ensemble Stream Modeling and Data-cleaning are sensor information processing systems have different training and testing methods by which their goals are cross-validated. This research examines a mechanism, which seeks to extract novel patterns by generating ensembles from data. The main goal of label-less stream processing is to process the sensed events to eliminate the noises that are uncorrelated, and choose the most likely model without over fitting thus obtaining higher model confidence. Higher quality streams can be realized by combining many short streams into an ensemble which has the desired quality. The framework for the investigation is an existing data mining tool. First, to accommodate feature extraction such as a bush or natural forest-fire event we make an assumption of the burnt area (BA*), sensed ground truth as our target variable obtained from logs. Even though this is an obvious model choice the results are disappointing. The reasons for this are two: One, the histogram of fire activity is highly skewed. Two, the measured sensor parameters are highly correlated. Since using non descriptive features does not yield good results, we resort to temporal features. By doing so we carefully eliminate the averaging effects; the resulting histogram is more satisfactory and conceptual knowledge is learned from sensor streams. Second is the process of feature induction by cross-validating attributes with single or multi-target variables to minimize training error. We use F-measure score, which combines precision and accuracy to determine the false alarm rate of fire events. The multi-target data-cleaning trees use information purity of the target leaf-nodes to learn higher order features. A sensitive variance measure such as f-test is performed during each node’s split to select the best attribute. Ensemble stream model approach proved to improve when using complicated features with a simpler tree classifier. The ensemble framework for data-cleaning and the enhancements to quantify quality of fitness (30% spatial, 10% temporal, and 90% mobility reduction) of sensor led to the formation of streams for sensor-enabled applications. Which further motivates the novelty of stream quality labeling and its importance in solving vast amounts of real-time mobile streams generated today.

This thesis deals with data extraction and data aggregation from heterogeneous web sources. The goal is to create a platform and a functional web application using appropriate technologies. The main focus of the thesis is on the application design and implementation. The application domain is accommodation or lease of apartments. For the data extraction, we use the portal API or a wrapper. Obtained data is stored in a document database. In this thesis, we managed to design and implement a system that allows to obtain rental ads from multiple web sources at the same time and to present them in a uniform way.

Gossip is a generic protocol which was designed for spreading information between nodes in large distributed decentralised systems. This protocol can be also used for many different applications including data aggregation, topology construction, etc. This work presents and describes a framework designed for facilitating modelling and simulation of Gossip-based systems.