Political contention has entered a new age. Over the past three years unprecedented large-scale movements have challenged states across the globe, and social media has been an important component in their development and articulation. With the advent of social media sites, such as Facebook and Twitter, ordinary people have the technological ability to instantaneously transcend space, time and resources (Aouraugh and Alexander 2011; Castells 2012; Earl and Kimport 2009, 2011; Eltantawy, Nahed and Wiest 2011; Gerbaudo 2012; Hands 2011; Holmes 2012; Mason 2012). Are we currently living in a historical moment where a new repertoire of contention is emerging? If so, how is social media changing the way we collectively contest for our interests? The theoretical framework I propose in this paper advances and elaborates a social geographic approach in the framing of political contention that emphasizes the importance of the spatiality and temporality created by the hashtag (#) in the development and articulation of today's social movements. In addition to secondary sources about the protests in Brazil (#VemPraRua), I draw on participant observations to analyze a new modular form of protest I call the "hashtag movement." I claim that the hashtag (#) creates a new space/time (Massey 1992, 2007; Soja 1996) that fundamentally shifts the process of nation-ness (Anderson 2006) and marks a new phase in the mediazation of modern culture (Thompson 1991); two fundamental shifts that I argue are comparable to the structural and cultural shifts that formed the modern repertoire of contention (Anderson 2006; Della Porta and Diani 1999; McAdam 1999; McAdam, Tarrow and Tilly 2001; Sewell 1990, 1996; Swidler 1986; Tarrow 1993, 1994; Tilly 1986, 1995a, 1995b; Young 2002). / text

Hate crime laws in both Canada and the United States purport to promote equality using the language of antidiscrimination law. National criminal codes in both countries authorize enhanced punishment for crimes motivated by “sexual orientation” but not “gender identity” or “gender expression.” Cities and states in the United States have also adopted hate crime laws, some of which denounce both homophobic and trans-phobic crimes. Hate crime penalty enhancement laws have been applied by courts in both Canada and the United States to establish a growing jurisprudence. In both countries, moreover, other hate crime laws contribute to official legal knowledge by regulating hate speech, hate crime statistics, and conduct equivalent to hate crimes in schools, workplaces, and elsewhere. Yet, despite the proliferation of hate crime laws and jurisprudence, governmental officials do not control all legal knowledge about hate crimes. Sociological “others” attend criminal sentencing proceedings and provide support to hate crime victims during prosecutions, but they also frame their own unofficial inquiries and announce their own classification decisions for hate-related events. In both Canada and the United States, nongovernmental groups contend both inside and outside official governmental channels to establish legal knowledge about homophobic and trans-phobic hate crimes. In two comparable Canadian and American cities, similar groups monitor and classify homophobic and trans-phobic attacks using a variety of information practices. Interviews with representatives of these groups reveal a relationship between the practices of each group and hate crime laws at each site. The results support one principal conclusion. The availability of local legislative power and a local mechanism for public review are key determinants of the sites and styles of nongovernmental contention about hate crimes. Where police gather and publish official hate crime statistics, the official classification system serves as both a site for mobilization, and a constraint on the styles of contention used by nongovernmental groups. Where police do not gather or publish hate crime statistics, nongovernmental groups are deprived of the resource represented by a local site for social contention, but their styles of contention are liberated from the subtle influences of an official hate crime classification system.

Les personnes présentant une déficience intellectuelle (DI) sont plus à risque de manifester des comportements problématiques dû, entre autres, à des déficits de communication, des syndromes génétiques, des problèmes de santé associés avec cette clientèle. Lorsque des personnes présentant une DI manifestent des comportements agressifs, la probabilité qu'elles soient blessées, isolées et qu'elles subissent des interventions plus restrictives est plus élevée. En outre, on observe une augmentation du risque de blessures chez leurs proches, une augmentation du stress des intervenants, ainsi qu'une augmentation des coûts reliés aux services dont la personne a besoin. On sait, par ailleurs, que des interventions adéquates et planifiées diminuent ces conséquences négatives pour la personne présentant une DI et son entourage. Le premier volet de cette thèse visait à identifier les variables personnelles, comportementales et environnementales reliées à l'utilisation des mesures de contrôle auprès des personnes présentant une déficience intellectuelle (DI) et manifestant des comportements agressifs. Pour ce faire, des questionnaires ont été envoyés dans les centres de réadaptation en déficience intellectuelle et trouble envahissant du développement (CRDITED) au Québec. Des données concernant 81 adultes présentant une DI et manifestant des comportements agressifs ont été obtenues. Les résultats révèlent que 63 % des participants ont vécu une mesure de contrôle dans les trois derniers mois. Plus précisément, 44,4 % ont subi une mesure d'isolement, 42 % une contention physique et 27,2 % une contention mécanique. De plus, le mode de communication, la prise d'anxiolytiques, la sévérité des comportements agressifs, la présence d'une analyse fonctionnelle au dossier ainsi que le fait d'avoir un intervenant ayant moins d'années d'expérience auprès de la clientèle se sont avérés des prédicteurs de l'utilisation d'une mesure de contrôle auprès des participants. Le second volet de cette thèse visait à documenter les perceptions des personnes présentant une DI et des intervenants envers l'utilisation des mesures de contrôle. Pour ce faire des entrevues ont été effectuées auprès de huit personnes présentant une DI et huit intervenants ayant participé au premier volet de cette thèse et ayant reçu ou appliqué une mesure de contrôle dans le dernier mois. Les entrevues portaient sur leur compréhension des mesures de contrôle, l'impact sur la relation, les émotions ressenties lors des interventions ainsi que les interventions alternatives possibles. Les résultats révèlent que les participants présentant une DI ainsi que les intervenants vivent beaucoup d'émotions négatives en lien avec les mesures de contrôle. Néanmoins, la majorité des participants semblent comprendre le but des mesures de contrôle et ont été en mesure d'identifier des interventions alternatives. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : déficience intellectuelle, comportements agressifs, mesures de contrôle, contention, isolement.

In this thesis we propose a cloud Input Shaper and Dynamic Resource Controller to provide application-level quality of service guarantees in cloud computing environments. The Input Shaper splits the cloud into two areas: one for shaped traffic that achieves quality of service targets, and one for overflow traffic that may not achieve the targets. The Dynamic Resource Controller profiles customers’ applications, then calculates and allocates the resources required by the applications to achieve given quality of service targets. The Input Shaper then shapes the rate of incoming requests to ensure that the applications achieve their quality of service targets based on the amount of allocated resources. To evaluate our system we create a new benchmark application that is suitable for use in cloud computing environments. It is designed to reflect the current design of cloud based applications and can dynamically scale each application tier to handle large and varying workload levels. In addition, the client emulator that drives the benchmark also mimics realistic user behaviors such as browsing from multiple tabs, using JavaScript, and has variable thinking and typing speeds. We show that a cloud management system evaluated using previous benchmarks could violate its estimated quality of service achievement rate by over 20%. The Input Shaper and Dynamic Resource Controller system consist of an application performance modeler, a resource allocator, decision engine, and an Apache HTTP server module to reshape the rate of incoming web requests. By dynamically allocating resources to applications, we show that their response times can be improved by as much as 30%. Also, the amount of resources required to host applications can be decreased by 20% while achieving quality of service objectives. The Input Shaper can reduce VMs’ resource utilization variances by 88%, and reduce the number of servers by 45%.

隨著行動通訊日漸普及，愈來愈多使用者希望能夠在任何時間、任何地點，不受限制地使用他們所需的服務。但是由於IEEE 802.11通訊協定的限制，導致使用者無法獲得適當的服務品質保障，特別是使用者在享受即時性的服務，例如：網路電話、視訊會議、線上即時影音等。雖然在2001年IEEE提出802.11e草案，提供具有服務品質的無線通訊協定，但是在服務品質的效能方面表現並非完美，特別是在隨時變化的真實網路中。我們提出一個隨時根據網路狀況動態調整系統參數且容易實作並相容於IEEE 802.11e網路下的新機制。根據實驗模擬結果中，我們可以看到我們所提出的方法提供比EDCA更好的服務品質，就語音方面的延遲而言，我們減少約80%，就視訊方面的延遲，我們改善約25%，而輸出量方面約增加45%的效能，並且提高系統約6% - 40%最大的容納數。而我們也導入了允入控管，根據事先模擬的結果及設定的系統參數門檻，達到允入控管的目地，確保系統所提供的服務品質能夠滿足使用者的需求。 / Wireless LANs (WLANs) based on the IEEE 802.11 family have recently become popular for allowing high data rates at relatively low cost. However, due to the limitation of 802.11 DCF, they are not suitable for real-time service and do not support service differentiation. To expand providing applications with quality of service (QoS) requirements in WLANs, the IEEE 802.11E Task Group was formed to enhance the current 802.11 medium access control (MAC) protocol. In real networks, the network condition is dynamic. The 802.11e cannot reach high performance of QoS. Besides, design of efficient MAC protocols with both high-throughput and low-jitter performance is the major focus in distributed contention-based MAC protocol research. In view of this, we propose a new and applicable approach, called Dynamic-EDCA (D-EDCA), adjusts the parameters of system based on the network condition. The result of　simulation shows that in comparison with EDCA, our proposed algorithm improves total throughput more than 45%, and the maximum of connection more than 40%. It also reduces voice delay about 80%, and video delay about 25%. Thus, the proposed D-EDCA utilizes bandwidth more efficiently and improves performance for service differentiation.

Hate crime laws in both Canada and the United States purport to promote equality using the language of antidiscrimination law. National criminal codes in both countries authorize enhanced punishment for crimes motivated by “sexual orientation” but not “gender identity” or “gender expression.” Cities and states in the United States have also adopted hate crime laws, some of which denounce both homophobic and trans-phobic crimes. Hate crime penalty enhancement laws have been applied by courts in both Canada and the United States to establish a growing jurisprudence. In both countries, moreover, other hate crime laws contribute to official legal knowledge by regulating hate speech, hate crime statistics, and conduct equivalent to hate crimes in schools, workplaces, and elsewhere. Yet, despite the proliferation of hate crime laws and jurisprudence, governmental officials do not control all legal knowledge about hate crimes. Sociological “others” attend criminal sentencing proceedings and provide support to hate crime victims during prosecutions, but they also frame their own unofficial inquiries and announce their own classification decisions for hate-related events. In both Canada and the United States, nongovernmental groups contend both inside and outside official governmental channels to establish legal knowledge about homophobic and trans-phobic hate crimes. In two comparable Canadian and American cities, similar groups monitor and classify homophobic and trans-phobic attacks using a variety of information practices. Interviews with representatives of these groups reveal a relationship between the practices of each group and hate crime laws at each site. The results support one principal conclusion. The availability of local legislative power and a local mechanism for public review are key determinants of the sites and styles of nongovernmental contention about hate crimes. Where police gather and publish official hate crime statistics, the official classification system serves as both a site for mobilization, and a constraint on the styles of contention used by nongovernmental groups. Where police do not gather or publish hate crime statistics, nongovernmental groups are deprived of the resource represented by a local site for social contention, but their styles of contention are liberated from the subtle influences of an official hate crime classification system.

[Truncated abstract] Parallel computing is recognised today as an important tool in the solution of a wide variety of computationally intensive problems, problems which were previously considered intractable. While it offers the promise of vastly increased performance, parallel computing introduces additional complexities which are not encountered with sequential processing. One of these is the scheduling problem, in which the individual tasks comprising a parallel program are scheduled onto the processors comprising the parallel architecture. The objective is to minimise execution time while still preserving the precedence relations between the tasks. Scheduling is of vital importance since a poor task schedule can undo any potential gains from the parallelism present in the application. Inappropriate scheduling can result in the hardware being used inefficiently, or worse, the program could run slower in parallel than on a single processor. The scheduling problem is one of the more difficult problems facing the parallel programmer. In fact, it is NP-complete in the general case. As a result, a large number of heuristic methods with sub-optimal performance but polynomial, rather than exponential, time complexity have been proposed. In order to simplify their algorithms, researchers have restricted the problem: by making assumptions concerning the parallel architecture or imposing limitations on the task graph representing the parallel program. The evolution of the task scheduling problem has involved the gradual relaxation of these restrictions. A major change occurred when the assumption of zero inter-processor communication costs was removed. This was driven by the increasing popularity of distributed-memory message-passing multiprocessors.

Submitted by Maria Medeiros (maria.dilva1@ufcg.edu.br) on 2018-08-06T15:07:04Z No. of bitstreams: 1 JOSÉ FERREIRA JÚNIOR - TESE (PPGCS) 2014.pdf: 3762511 bytes, checksum: 7f275604704f424efd04b446e18ec2f9 (MD5) / Made available in DSpace on 2018-08-06T15:07:04Z (GMT). No. of bitstreams: 1 JOSÉ FERREIRA JÚNIOR - TESE (PPGCS) 2014.pdf: 3762511 bytes, checksum: 7f275604704f424efd04b446e18ec2f9 (MD5) Previous issue date: 2014 / O objetivo desta pesquisa foi investigar como, em Serra Talhada e em Triunfo, cidades da Mesorregião do Sertão Pernambucano, Microrregião do Pajeú, verifica-se o uso da memória lampiônica pelos produtores culturais e, no referente a esse uso, como são construídos os discursos desses atores sociais. Procedeu-se investigação bibliográfica. Foram examinadas obras diversificadas que tratam do assunto memória e mercadorização da memória. Também foram consultadas literaturas que circulam no âmbito regional, cujos autores são produtores culturais serratalhadenses e triunfenses e ex-soldados de volantes. Concernente à teoria empregada, buscou-se embasamento no discurso weberiano da ação social, mais especificamente, a que busca uma finalidade específica. Somado a essa incursão bibliográfica, também foram feitas entrevistas, tendo como instrumento de obtenção de dados o questionário semi-estruturado. Os conteúdos dos discursos dos sujeitos entrevistados foram analisados e utilizados na construção do texto que compõe esta tese. Como resultados, percebeu-se, de fato, a existência de disputa pela memória citada e, em decorrência de tal procedimento, a constatação de que, embora a intencionalidade seja a mesma, verifica-se a utilização de estratégias diferentes para se obter benefício da memória em questão. Enquanto em Serra Talhada a memória lampiônica é diretamente mercadejada, satisfazendo uma demanda turística, em Triunfo, seu uso se faz quando é agregada aos atrativos da cidade, ou seja, no espaço triunfense, a memória de Lampião funciona como elemento agregador de valor àquilo que a cidade, por si só, já oferece ao turista. / The objective of this research was to investigate how, Serra Talhada and Triunfo , the cities of Meso Hinterland Pernambucano microrregion the Pajeú, there is the use of memory lampiônica by cultural producers, and in respect of such use , how they are constructed discourses these social actors. Proceeded bibliographic research . Diversified works dealing with the subject memory and commodification of memory were examined. Literatures circulating regionally, whose authors are serratalhadenses cultural producers and triunfenses and former soldiers flywheels were also consulted. Concerning the theory employed, we sought grounding in the Weberian discourse of social action, more specifically, to search a particular purpose. Added to this literature raid, interviews were also carried out, and as a tool for obtaining data semi - structured questionnaire. The contents of the interviews subjects were analyzed and used in the construction of the text that makes up this thesis. As a result, it was realized, in fact, the existence of the dispute quoted memory and, as a result of such a procedure, the finding that, although the intention is the same, the use of different strategies it appears to give the benefit of the memory in question. While in Serra Talhada hewn lampiônica the memory is directly marketed, satisfying a demand for tourism, in Triunfo its use becomes attractive when it is aggregated to the city , or in triunfense space, memory of Lampião works as aggregator element value to what the city itself already offers the tourist

Cette recherche doctorale aborde le travail de sept organisations de mémoire et de droits humains de Santiago, au Chili, et les luttes dans lesquelles elles sont engagées. Le passé et sa représentation dans l'espace public sont une source de conflit depuis la fin du régime Pinochet. D'allure spectrale, ces sujets réapparaissent dans le présent sous plusieurs formes alors que les organisations de mémoire et de droits humains entrent en lutte relativement aux sens et à l'importance que doit prendre ce passé, de même qu’à la légitimité devant être accordée aux acteurs mobilisant ce passé dans l'arène publique. Or, ces luttes et conflits se produisent dans des dynamiques sociopolitiques particulières et engagent une multitude d'acteurs ayant des représentations différenciées du passé. Cet état de fait appelle à une reformulation de la conceptualisation du rôle des acteurs collectifs afin de mettre en exergue ces différences dans la manière de représenter le passé, certes, mais également dans les stratégies analogues et opposées de le faire. Cette thèse entreprend, de manière diachronique, la reconstruction des principaux débats s'étant produits autour des questions de l'impunité, de la vérité et de la mémoire. Couvrant la période de 1998 à 2018, elle débute avec la détention d'Augusto Pinochet à Londres et se termine avec la fin du second mandat de la première femme présidente, Michelle Bachelet. Les sept organisations étudiées vont de regroupements de familles et de survivants à des sites de mémoire et de conscience. Par le biais des discours produits par ces organisations, j'examine des périodes historiques particulières en me concentrant sur des moments forts de ces conjonctures afin de mettre en évidence les luttes de sens et les conflits de légitimité relatifs à ces époques. De manière concrète, une attention particulière est portée à l'analyse des principaux sujets de contentieux, la façon dont les acteurs se mobilisent, leurs objectifs et comment le passé est évoqué dans l'espace public. L'une des fins de cette thèse est de mettre en exergue la relation conflictuelle entre les organisations et les administrations post-dictature. Cette dimension permet ainsi de comprendre le malaise des représentations politiques transparaissant en filigrane des interventions publiques des porteurs de mémoire. À partir d'un cadre théorique original fondé à la fois dans la sociologie de la mémoire, la sociologie des mouvements sociaux et l'analyse de discours, cette thèse qualitative met à profit un corpus de documents diversifié, notamment des journaux nationaux, des déclarations publiques, de même que des textes provenant de la plate-forme web des organisations étudiées. Ces documents sont examinés et analysés afin de mettre en évidence les représentations et les dynamiques qui les soulignent. S'inspirant de la méthode employée par Passy et Giugni (2005 : 903), ce travail étudie le corpus de documents recueillis afin de faire la lumière sur « qui intervient (les acteurs), comment (les répertoires d'action) et dans quel but (contenu des revendications) » (Passy & Giugni, 2005: 903). Cette technique permet de retracer les acteurs, leur répertoire d'action, leurs représentations du passé, la terminologie qu'ils utilisent et les traces de leurs interactions avec les autres organisations ou représentations opposées du passé.

abstract: With the massive multithreading execution feature, graphics processing units (GPUs) have been widely deployed to accelerate general-purpose parallel workloads (GPGPUs). However, using GPUs to accelerate computation does not always gain good performance improvement. This is mainly due to three inefficiencies in modern GPU and system architectures. First, not all parallel threads have a uniform amount of workload to fully utilize GPU’s computation ability, leading to a sub-optimal performance problem, called warp criticality. To mitigate the degree of warp criticality, I propose a Criticality-Aware Warp Acceleration mechanism, called CAWA. CAWA predicts and accelerates the critical warp execution by allocating larger execution time slices and additional cache resources to the critical warp. The evaluation result shows that with CAWA, GPUs can achieve an average of 1.23x speedup. Second, the shared cache storage in GPUs is often insufficient to accommodate demands of the large number of concurrent threads. As a result, cache thrashing is commonly experienced in GPU’s cache memories, particularly in the L1 data caches. To alleviate the cache contention and thrashing problem, I develop an instruction aware Control Loop Based Adaptive Bypassing algorithm, called Ctrl-C. Ctrl-C learns the cache reuse behavior and bypasses a portion of memory requests with the help of feedback control loops. The evaluation result shows that Ctrl-C can effectively improve cache utilization in GPUs and achieve an average of 1.42x speedup for cache sensitive GPGPU workloads. Finally, GPU workloads and the co-located processes running on the host chip multiprocessor (CMP) in a heterogeneous system setup can contend for memory resources in multiple levels, resulting in significant performance degradation. To maximize the system throughput and balance the performance degradation of all co-located applications, I design a scalable performance degradation predictor specifically for heterogeneous systems, called HeteroPDP. HeteroPDP predicts the application execution time and schedules OpenCL workloads to run on different devices based on the optimization goal. The evaluation result shows HeteroPDP can improve the system fairness from 24% to 65% when an OpenCL application is co-located with other processes, and gain an additional 50% speedup compared with always offloading the OpenCL workload to GPUs. In summary, this dissertation aims to provide insights for the future microarchitecture and system architecture designs by identifying, analyzing, and addressing three critical performance problems in modern GPUs. / Dissertation/Thesis / Doctoral Dissertation Computer Engineering 2017