Managed Acceleration for In-Memory Database Analytic Workloads

In-Memory Databases (IMDBs), such as SAP HANA, enable new levels of database performance by removing the disk bottleneck and by compressing data in memory. The consequence of this improved performance means that reports and analytic queries can now be processed on demand. Therefore, the goal is now to provide near real-time responses to compute and data intensive analytic queries. To facilitate this, much work has investigated the use of acceleration technologies within the database context. While current research into the application of these technologies has yielded positive results, they have tended to focus on single database tasks or on isolated single user requests. This paper uses SHEPARD, a framework for managing accelerated tasks across shared heterogeneous resources, to introduce acceleration into an IMDB. Results show how, using SHEPARD, multiple simultaneous user queries all receive speed-up by using a shared pool of accelerators. Results also show that offloading analytic tasks onto accelerators can have indirect benefits for other database workloads by reducing contention for CPU resources.

Context-based wireless mesh networks

In the modern society, new devices, applications and technologies, with sophisticated capabilities, are converging in the same network infrastructure. Users are also increasingly demanding in personal preferences and expectations, desiring Internet connectivity anytime and everywhere. These aspects have triggered many research efforts, since the current Internet is reaching a breaking point trying to provide enough flexibility for users and profits for operators, while dealing with the complex requirements raised by the recent evolution. Fully aligned with the future Internet research, many solutions have been proposed to enhance the current Internet-based architectures and protocols, in order to become context-aware, that is, to be dynamically adapted to the change of the information characterizing any network entity. In this sense, the presented Thesis proposes a new architecture that allows to create several networks with different characteristics according to their context, on the top of a single Wireless Mesh Network (WMN), which infrastructure and protocols are very flexible and self-adaptable. More specifically, this Thesis models the context of users, which can span from their security, cost and mobility preferences, devices’ capabilities or services’ quality requirements, in order to turn a WMN into a set of logical networks. Each logical network is configured to meet a set of user context needs (for instance, support of high mobility and low security). To implement this user-centric architecture, this Thesis uses the network virtualization, which has often been advocated as a mean to deploy independent network architectures and services towards the future Internet, while allowing a dynamic resource management. This way, network virtualization can allow a flexible and programmable configuration of a WMN, in order to be shared by multiple logical networks (or virtual networks - VNs). Moreover, the high level of isolation introduced by network virtualization can be used to differentiate the protocols and mechanisms of each context-aware VN. This architecture raises several challenges to control and manage the VNs on-demand, in response to user and WMN dynamics. In this context, we target the mechanisms to: (i) discover and select the VN to assign to an user; (ii) create, adapt and remove the VN topologies and routes. We also explore how the rate of variation of the user context requirements can be considered to improve the performance and reduce the complexity of the VN control and management. Finally, due to the scalability limitations of centralized control solutions, we propose a mechanism to distribute the control functionalities along the architectural entities, which can cooperate to control and manage the VNs in a distributed way.

Mining biomedical information from scientific literature

The rapid evolution and proliferation of a world-wide computerized network, the Internet, resulted in an overwhelming and constantly growing amount of publicly available data and information, a fact that was also verified in biomedicine. However, the lack of structure of textual data inhibits its direct processing by computational solutions. Information extraction is the task of text mining that intends to automatically collect information from unstructured text data sources. The goal of the work described in this thesis was to build innovative solutions for biomedical information extraction from scientific literature, through the development of simple software artifacts for developers and biocurators, delivering more accurate, usable and faster results. We started by tackling named entity recognition - a crucial initial task - with the development of Gimli, a machine-learning-based solution that follows an incremental approach to optimize extracted linguistic characteristics for each concept type. Afterwards, Totum was built to harmonize concept names provided by heterogeneous systems, delivering a robust solution with improved performance results. Such approach takes advantage of heterogenous corpora to deliver cross-corpus harmonization that is not constrained to specific characteristics. Since previous solutions do not provide links to knowledge bases, Neji was built to streamline the development of complex and custom solutions for biomedical concept name recognition and normalization. This was achieved through a modular and flexible framework focused on speed and performance, integrating a large amount of processing modules optimized for the biomedical domain. To offer on-demand heterogenous biomedical concept identification, we developed BeCAS, a web application, service and widget. We also tackled relation mining by developing TrigNER, a machine-learning-based solution for biomedical event trigger recognition, which applies an automatic algorithm to obtain the best linguistic features and model parameters for each event type. Finally, in order to assist biocurators, Egas was developed to support rapid, interactive and real-time collaborative curation of biomedical documents, through manual and automatic in-line annotation of concepts and relations. Overall, the research work presented in this thesis contributed to a more accurate update of current biomedical knowledge bases, towards improved hypothesis generation and knowledge discovery.

Network virtualisation from an operator perspective

Network virtualisation is seen as a promising approach to overcome the so-called “Internet impasse” and bring innovation back into the Internet, by allowing easier migration towards novel networking approaches as well as the coexistence of complementary network architectures on a shared infrastructure in a commercial context. Recently, the interest from the operators and mainstream industry in network virtualisation has grown quite significantly, as the potential benefits of virtualisation became clearer, both from an economical and an operational point of view. In the beginning, the concept has been mainly a research topic and has been materialized in small-scale testbeds and research network environments. This PhD Thesis aims to provide the network operator with a set of mechanisms and algorithms capable of managing and controlling virtual networks. To this end, we propose a framework that aims to allocate, monitor and control virtual resources in a centralized and efficient manner. In order to analyse the performance of the framework, we performed the implementation and evaluation on a small-scale testbed. To enable the operator to make an efficient allocation, in real-time, and on-demand, of virtual networks onto the substrate network, it is proposed a heuristic algorithm to perform the virtual network mapping. For the network operator to obtain the highest profit of the physical network, it is also proposed a mathematical formulation that aims to maximize the number of allocated virtual networks onto the physical network. Since the power consumption of the physical network is very significant in the operating costs, it is important to make the allocation of virtual networks in fewer physical resources and onto physical resources already active. To address this challenge, we propose a mathematical formulation that aims to minimize the energy consumption of the physical network without affecting the efficiency of the allocation of virtual networks. To minimize fragmentation of the physical network while increasing the revenue of the operator, it is extended the initial formulation to contemplate the re-optimization of previously mapped virtual networks, so that the operator has a better use of its physical infrastructure. It is also necessary to address the migration of virtual networks, either for reasons of load balancing or for reasons of imminent failure of physical resources, without affecting the proper functioning of the virtual network. To this end, we propose a method based on cloning techniques to perform the migration of virtual networks across the physical infrastructure, transparently, and without affecting the virtual network. In order to assess the resilience of virtual networks to physical network failures, while obtaining the optimal solution for the migration of virtual networks in case of imminent failure of physical resources, the mathematical formulation is extended to minimize the number of nodes migrated and the relocation of virtual links. In comparison with our optimization proposals, we found out that existing heuristics for mapping virtual networks have a poor performance. We also found that it is possible to minimize the energy consumption without penalizing the efficient allocation. By applying the re-optimization on the virtual networks, it has been shown that it is possible to obtain more free resources as well as having the physical resources better balanced. Finally, it was shown that virtual networks are quite resilient to failures on the physical network.

You Can't Always Get What You Want: A Gewirthian Model of Rational Autonomy in Abortion

The law regulating the availability of abortion is problematic both legally and morally. It is dogmatic in its requirements of women and doctors and ignorant of would-be fathers. Practically, its usage is liberal - with s1(1)(a) Abortion Act 1967 treated as a ‘catch all’ ground - it allows abortion on demand. Yet this is not reflected in the ‘law’. Against this outdated legislation I propose a model of autonomy which seeks to tether our moral concerns with a new legal approach to abortion. I do so by maintaining that a legal conception of autonomy is derivable from the categorical imperative resulting from Gewirth’s argument to the Principle of Generic Consistency: Act in accordance with the generic rights of your recipients as well as of yourself. This model of Gewirthian Rational Autonomy, I suggest, provides a guide for both public and private notions of autonomy and how our autonomous interests can be balanced across social structures in order to legitimately empower choice. I claim, ultimately, that relevant rights in the context of abortion are derivable from this model.

Explorador integrado de serviços de armazenamento na cloud: cloud cockpit

Projeto para obtenção do grau de Mestre em Engenharia Informática e de Computadores

A high-rate mimo receiver in an FPGA

This paper describes the hardware implementation of a High-Rate MIMO Receiver in an FPGA for three modulations, namely BPSK, QPSK and 16-QAM based on the Alamouti scheme. The implementation with 16-QAM achieves more than 1.6 Gbps with 66% of the resources of a medium-sized Virtex-4 FPGA. This results indicate that the Alamouti scheme is a good design option for hardware implementation of a high-rate MIMO receiver. Also, using an FPGA, the modulation can be dynamically changed on demand.

Preadjunct questions as a learning strategy for older adults

This study investigated the effectiveness of comprehension level preadjunct questions as a learning strategy for older adults in a classroom setting. Fifty-five adults from 55 to 70 years of age were randomly assigned to two groups, the preadjunct question group and a no-question control group. They viewed a video on high blood pressure and completed a recall posttest immediately after viewing the video and again seven days tater. Results demonstrated that there was no significant difference between groups. However, the no-question control group obtained a higher mean score on both the immediate and delayed recall tests than did the preadjunct question group. Nevertheless, significant differences in posttest scores were found related to educational levels and prior knowledge about high blood pressure. Results obtained were explained in terms of resource theory of cognitive aging.

Niagara Suspension Bridge Bank note, 1841

The Niagara Suspension Bridge Bank operated in Queenston in 1840. The bank issued notes in denominations of ten dollars, five dollars and one dollar, and featured a drawing of the Queenston-Lewiston Bridge, ten years prior to its construction. The notes are signed by the bank’s Cashier, Gilbert McMicken, and President, Joseph Hamilton. The bank failed a year after its establishment.

Modèle de confiance et ontologie probabiliste pilotés par réseaux bayésiens pour la gestion des accords de services dans l’environnement de services infonuagiques

L’infonuage est un nouveau paradigme de services informatiques disponibles à la demande qui a connu une croissance fulgurante au cours de ces dix dernières années. Le fournisseur du modèle de déploiement public des services infonuagiques décrit le service à fournir, le prix, les pénalités en cas de violation des spécifications à travers un document. Ce document s’appelle le contrat de niveau de service (SLA). La signature de ce contrat par le client et le fournisseur scelle la garantie de la qualité de service à recevoir. Ceci impose au fournisseur de gérer efficacement ses ressources afin de respecter ses engagements. Malheureusement, la violation des spécifications du SLA se révèle courante, généralement en raison de l’incertitude sur le comportement du client qui peut produire un nombre variable de requêtes vu que les ressources lui semblent illimitées. Ce comportement peut, dans un premier temps, avoir un impact direct sur la disponibilité du service. Dans un second temps, des violations à répétition risquent d'influer sur le niveau de confiance du fournisseur et sur sa réputation à respecter ses engagements. Pour faire face à ces problèmes, nous avons proposé un cadre d’applications piloté par réseau bayésien qui permet, premièrement, de classifier les fournisseurs dans un répertoire en fonction de leur niveau de confiance. Celui-ci peut être géré par une entité tierce. Un client va choisir un fournisseur dans ce répertoire avant de commencer à négocier le SLA. Deuxièmement, nous avons développé une ontologie probabiliste basée sur un réseau bayésien à entités multiples pouvant tenir compte de l’incertitude et anticiper les violations par inférence. Cette ontologie permet de faire des prédictions afin de prévenir des violations en se basant sur les données historiques comme base de connaissances. Les résultats obtenus montrent l’efficacité de l’ontologie probabiliste pour la prédiction de violation dans l’ensemble des paramètres SLA appliqués dans un environnement infonuagique.

Asymétries locomotrices réelles et perçues lors de différentes conditions de marche sur tapis roulant chez les personnes saines et hémiparétiques

L'asymétrie locomotrice chez les personnes hémiparétiques à la suite d'un accident vasculaire cérébral (AVC) est fréquemment observée en clinique. Les causes sous-jacentes à cette asymétrie et sa persistance malgré les interventions sont source de nombreuses interrogations. Ce projet avait donc pour objectif de développer de nouvelles connaissances sur les facteurs pouvant expliquer l'asymétrie locomotrice après un AVC. Les objectifs principaux étaient de 1) quantifier les moments et les niveaux d'effort musculaire des membres inférieurs lors de différentes conditions de marche sur le tapis roulant à double courroie (DC) afin de déterminer si ceux-ci sont explicatifs de l'asymétrie locomotrice, et de 2) quantifier la capacité de perception de l'asymétrie locomotrice des personnes saines et hémiparétiques. L’hypothèse générale était que l’asymétrie locomotrice est une stratégie utilisée par les personnes hémiparétiques pour symétriser ou limiter les niveaux d’effort musculaire, tels qu’évalués par le taux d'utilisation musculaire mécanique (TUM). Les résultats ont été présentés dans quatre articles (#2 à #5). La première étude a quantifié l'effet d'un protocole sur tapis roulant à DC visant à modifier l'asymétrie spatio-temporelle sur les moments musculaires aux membres inférieurs des personnes saines et hémiparétiques. Globalement, les résultats ont démontré que placer le membre inférieur parétique sur la courroie lente augmentait le moment fléchisseur plantaire (FP) parétique en post-adaptation et que cette augmentation était corrélée avec l'augmentation de la longueur de pas controlatérale. La deuxième étude a démontré que les personnes présentant de hauts niveaux d'effort dans les FP parétiques avaient une asymétrie des efforts dans leurs FP et extenseurs de hanche (EH) alors que le groupe avec de bas niveaux d'effort dans les FP avait des efforts symétriques. De plus, les deux groupes présentaient des moments FP asymétriques, mais cette asymétrie était reliée à l'asymétrie de forces résiduelles seulement chez ceux présentant de hauts niveaux d'effort. Cela suggère que la faiblesse musculaire joue un rôle important dans la réduction du moment FP parétique chez ceux avec de hauts niveaux d'effort alors que d'autres facteurs doivent être considérés pour expliquer l'asymétrie des moments à la marche des personnes avec de bas niveaux d'effort. La troisième étude a démontré que la symétrisation du patron de marche (temporelle ou spatiale) amenait des efforts musculaires élevés dans les FP parétiques, atteignant des niveaux moyens supérieurs à 85 %. La quatrième étude a évalué la capacité de perception de l'asymétrie locomotrice des personnes saines avec un protocole sur tapis roulant à DC et a démontré qu'elles percevaient leur patron de marche comme asymétrique lorsque leur ratio d'asymétrie temporelle atteignait le seuil de 1,05, avec une variabilité intra-individu faible, suggérant une bonne capacité de ces personnes à détecter l'asymétrie. Les résultats supplémentaires de cette thèse ont démontré que les personnes hémiparétiques percevaient mieux l'asymétrie temporelle que spatiale. En somme, ces résultats suggèrent que la force musculaire et les efforts musculaires des FP parétiques sont à considérer comme des facteurs explicatifs de l'asymétrie locomotrice des personnes hémiparétiques. D'autres études sont nécessaires pour statuer sur les capacités de perception de l'asymétrie locomotrice des personnes hémiparétiques.

Intervention permettant une meilleure prise en charge des autosoins par la clientèle adulte de chirurgie intestinale

Rapport de stage présenté à la Faculté des études supérieures en vue de l'obtention du grade de Maitre és sciences (M.Sc) en sciences infirmières option expertise-conseil en soins infirmiers

Basic laser printing WSA (Mac)

Video on how to print to a Ricoh laser printer at WSA from a Mac, applies University wide just don't print WSA Colour or WSA Black & White if you are not WSA based.

Making a PDF (Photoshop)

Video on how to create a basic PDF for standard paper based Printing using Photoshop. Ideal format for all printers worldwide.

How to setup a basic Photoshop with Print in view (300dpi)

Video on how to setup a Basic Photoshop document with the view to outputting a Printable PDF.