On a qualifié de « migration des images » le déplacement des images en mouvement depuis les salles de projection jusque dans les salles des musées. Cette « migration » a d’abord constitué le socle initial de ce travail, en fixant les pourtours de son contexte : les images en mouvement, le musée et ses collections, leur dialogue depuis le milieu des années 1990. Délaissant ce sens premier de l’expression, cette étude met au travail son sens figuré : l’idée de migration convie une économie, un système d’échanges, de circulation, un type d’activité propre aux images en mouvement qui élargit ce contexte de départ. La force de ces images se loge dans un geste de décentrement. Leur mobilité les empêche d’accéder à un statut ou à une position dominante. Une résistance à la saisie globale accompagne ces images qui préfèrent les interstices, l’entre-deux, les failles, voire une certaine forme de clandestinité : elles échappent aux normes. Le décentrement qualifiera ici une méthode d’analyse, le choix d’un corpus qui va des années 1970 à nos jours, et la construction générale de cette recherche. Il permet de mesurer des distances, de saisir les sinuosités, les accidents, qui ponctuent l’histoire des images en mouvement. Point de départ : ce travail s’arrime au décentrement historique constitué par la rupture d’avec le discours moderniste dans les années 1970. Les œuvres contemporaines réunies héritent de cette éviction et la discutent. Les trois chapitres, thématiques, commentent ce décentrement qui bouscule les principes muséaux incarnés par l’institution du white cube. L’archive, le document et le jeu : trois façons de contourner les réflexes de l’essentialisme. Fondé sur des études d’œuvres, qui sont autant d’études de cas, accueillant sur un plan qui se veut tabulaire les images, les textes et les films, ce travail rend compte du savoir, étoilé et ramifié, que dispensent, de façon si singulière, les images en mouvement. / The dislocation of images in motion from screening rooms to museum rooms has been addressed in terms of a « migration of images ». This « migration » was the initial foundation of the present work, determining its perimeter: images in motion, the museum and its collections, the dialogue between them from the mid-1990s onwards. Leaving behind the literal meaning of the expression, this study explores its fi gurative senses: the notion of « migration » calls for an economy, a system of exchange and circulation, an activity specific to images in motion that broadens its initial context. The power of images in motion relies upon a gesture of decentering. Th eir mobility prevents them from reaching a dominant position or attaining an important status. Th ese images resist to global analysis, preferring the interstices, the in-between, the faults, and what resembles to a form of underground activity instead. In other words, they resist the standards. Th e notion of « decentering » describes here three different dimensions: an analytical method, the constitution of a corpus ranging from the 1970s to the present-day, and the general conception of this research. Th e concept allows for the measuring of distances, the grasping of certain infl exions, or accidents, which rhythm the history of images in motion. Our starting point is the historical « decentering » illustrated by the break with modernist discourse in the 1970s. Th e contemporary works that we discuss inherit this eviction and comment on it. Our three thematic chapters comment this shift that challenges the museological principles embodied by the whitecube. Th e archive, the document and staging : three ways of circumventing the refl exes of essentialism. Based upon the close study of works, conceived as case-studies, accommodating for images, texts and films within a tabular plan, this work accounts for the singular knowledge provided by images in motion.

The Semantic Web is a vision to create a “web of knowledge”; an extension of the Web as we know it which will create an information space which will be usable by machines in very rich ways. The technologies which make up the Semantic Web allow machines to reason across information gathered from the Web, presenting only relevant results and inferences to the user. Users of the Web in its current form assess the credibility of the information they gather in a number of different ways. If processing happens without the user being able to check the source and credibility of each piece of information used in the processing, the user must be able to trust that the machine has used trustworthy information at each step of the processing. The machine should therefore be able to automatically assess the credibility of each piece of information it gathers from the Web. A case study on advanced checks for website credibility is presented, and the site presented in the case presented is found to be credible, despite failing many of the checks which are presented. A website with a backend based on RDF technologies is constructed. A better understanding of RDF technologies and good knowledge of the RAP and Redland RDF application frameworks is gained. The second aim of constructing the website was to gather information to be used for testing various trust metrics. The website did not gain widespread support, and therefore not enough data was gathered for this. Techniques for presenting RDF data to users were also developed during website development, and these are discussed. Experiences in gathering RDF data are presented next. A scutter was successfully developed, and the data smushed to create a database where uniquely identifiable objects were linked, even where gathered from different sources. Finally, the use of digital signature as a means of linking an author and content produced by that author is presented. RDF/XML canonicalisation is discussed in the provision of ideal cryptographic checking of RDF graphs, rather than simply checking at the document level. The notion of canonicalisation on the semantic, structural and syntactic levels is proposed. A combination of an existing canonicalisation algorithm and a restricted RDF/XML dialect is presented as a solution to the RDF/XML canonicalisation problem. We conclude that a trusted Semantic Web is possible, with buy in from publishing and consuming parties.

Enterprise document management has been in a long-term progress. Not long ago only files and folders were enough for enterprise document management, but nowadays most of enterprise documents are in the electronic form. The electronisation of enterprise documents has carried simplified manipulation, but it has also caused disorderliness in many cases. Systems called Document managament system (DMS) bring support for better management of electronic enterprise documents. This diploma thesis deals with the implementation of DMS into the environment of a specific company where it is designed to support effective function concerning enterprise document management.

Cette thèse explore l'usage des Grammaires Categorielles Abstraites (CGA) pour la Génération Automatique de Texte (GAT) dans un contexte industriel. Les systèmes GAT basés sur des théories linguistiques ont un long historique, cependant ils sont relativement peu utilisés en industrie, qui préfère les approches plus "pragmatiques", le plus souvent pour des raisons de simplicité et de performance. Cette étude montre que les avancées récentes en linguistique computationnelle permettent de concilier le besoin de rigueur théorique avec le besoin de performance, en utilisant CGA pour construire les principaux modules d'un système GAT de qualité industrielle ayant des performances comparables aux méthodes habituellement utilisées en industrie. / This thesis explores the usage of Abstract Categorial Grammars (ACG) for Natural Language Generation (NLG) in an industrial context. While NLG system based on linguistic theories have a long history, they are not prominent in industry, which, for the sake of simplicity and efficiency, usually prefer more ``pragmatic" methods. This study shows that recent advances in computational linguistics allow to conciliate the requirements of soundness and efficiency, by using ACG to build the main elements of a production grade NLG framework (document planner and microplanner), with performance comparable to existing, less advanced methods used in industry

Bakgrund: Vägen mot rättigheter för personer med intellektuell funktionsnedsättning (ID) har varit lång. Ännu idag krävs fortsatta ansträngningar då personer med ID har avsevärt sämre hälsa och medellivslängden är minst 20 år kortare. Detta är undvikbart och ojämlikt samt en överträdelse mot lagar och förordningar. Ett paradigmskifte krävs för att undanröja hinder för tillgänglig och jämlik hälso- och sjukvård för personer med ID. Syfte: Syftet var att undersöka hur jämlikhet och tillgänglighet för personer med intellektuella funktionsnedsättningar synliggörs i hälso-och sjukvårdsregioners styrdokument. Metod: Dokumentgranskning med innehållsanalys av kombinerad typ, såväl kvalitativ som kvantitativ. Resultat: Många regioner berör i sina styrdokument flera aspekter av tillgänglig och jämlik hälso- och sjukvård för personer med ID. Dock är det enbart i få dokument som åtgärder, uppföljningar och mätbara mål konkretiseras. Flera regioner nämner mänskliga rättigheter och kopplar även ihop detta med begreppet Känsla Av Sammanhang (KASAM). Konklusion: Rätt utformade och konkreta styrdokument, tillsammans med organisatoriska anpassningar verkar ge goda förutsättningar för såväl synliggörande av målgruppen samt implementering av åtgärder för att gynna tillgänglighet och jämlikhet. / Background: The efforts towards equality for persons with intellectual disabilities(ID) has been under way for a long time. Still today, continuous efforts are needed when persons with ID suffer from significantly worse health and with a life expectancy at least twenty years shorter than the rest of the population. The situation is avoidable and unequal as well as a violation against laws and regulations. A paradigm shift is needed to remove obstacles for accessable and equal healthcare for persons with ID. Aim: The aim of the study was to examine how equality and accessibility for persons with intellectual disablities is made visible in municipal health care policy documents. Method: Document analysis with mixed method content analysis, qualitative and quantitative. Results: Policy documents in several municipal healthcare reflect on several aspects of accessible and equal healthcare for persons with ID. However, only a few of them mention concrete measures, follow-ups and measurable goals. Several regions mention human rights and relate it to the concept of sence of coherence (SOC). Conclusion: Adequately formulated and concrete policy documents, combined with organizational adaptations seem to facilitate visibility for persons with ID andfurthermore the implementation of measures to promote accessibility and equality.

Indiana University-Purdue University Indianapolis (IUPUI) / Physicians receive multiple CCDs for a single patient encompassing various encounters and medical history recorded in different information systems. It is cumbersome for providers to explore different pages of CCDs to find specific data which can be duplicated or even conflicted. This study describes the steps towards a system that integrates multiple CCDs into one consolidated document for viewing or processing patient-level data. Also, the impact of the system on healthcare providers’ perceived workload is evaluated. A modular system is developed to consolidate and de-duplicate CDA-based documents. The system is engineered to be scalable, extensible and open source. The system’s performance and output has evaluated first based on synthesized data and later based on real-world CCDs obtained from INPC database. The accuracy of the consolidation system along with the gaps in identification of the duplications were assessed. Finally, the impact of the system on healthcare providers’ workload is evaluated using NASA TLX tool. All of the synthesized CCDs were successfully consolidated, and no data were lost. The de-duplication accuracy was 100% based on synthesized data and the processing time for each document was 1.12 seconds. For real-world CCDs, our system de-duplicated 99.1% of the problems, 87.0% of allergies, and 91.7% of medications. Although the accuracy of the system is still very promising, however, there is a minor inaccuracy. Due to system improvements, the processing time for each document is reduced to average 0.38 seconds for each CCD. The result of NASA TLX evaluation shows that the system significantly decreases healthcare providers’ perceived workload. Also, it is observed that information reconciliation reduces the medical errors. The time for review of medical documents review time is significantly reduced after CCD consolidation. Given increasing adoption and use of Health Information Exchange (HIE) to share data and information across the care continuum, duplication of information is inevitable. A novel system designed to support automated consolidation and de-duplication of information across clinical documents as they are exchanged shows promise. Future work is needed to expand the capabilities of the system and further test it using heterogeneous vocabularies across multiple HIE scenarios.

No description available.

Throughout the history, humans continue to generate an ever-growing volume of documents about a wide range of topics. We now rely on computer programs to automatically process these vast collections of documents in various applications. Many applications require a quantitative measure of the document similarity. Traditional methods first learn a vector representation for each document using a large corpus, and then compute the distance between two document vectors as the document similarity.In contrast to this corpus-based approach, we propose a straightforward model that directly discovers the topics of a document by clustering its words, without the need of a corpus. We define a vector representation called normalized bag-of-topic-embeddings (nBTE) to encapsulate these discovered topics and compute the soft cosine similarity between two nBTE vectors as the document similarity. In addition, we propose a logistic word importance function that assigns words different importance weights based on their relative discriminating power.Our model is efficient in terms of the average time complexity. The nBTE representation is also interpretable as it allows for topic discovery of the document. On three labeled public data sets, our model achieved comparable k-nearest neighbor classification accuracy with five stateof-art baseline models. Furthermore, from these three data sets, we derived four multi-topic data sets where each label refers to a set of topics. Our model consistently outperforms the state-of-art baseline models by a large margin on these four challenging multi-topic data sets. These works together provide answers to the research question of this thesis:Can we construct an interpretable document represen-tation by clustering the words in a document, and effectively and efficiently estimate the document similarity? / Under hela historien fortsätter människor att skapa en växande mängd dokument om ett brett spektrum av publikationer. Vi förlitar oss nu på dataprogram för att automatiskt bearbeta dessa stora samlingar av dokument i olika applikationer. Många applikationer kräver en kvantitativmått av dokumentets likhet. Traditionella metoder först lära en vektorrepresentation för varje dokument med hjälp av en stor corpus och beräkna sedan avståndet mellan two document vektorer som dokumentets likhet.Till skillnad från detta corpusbaserade tillvägagångssätt, föreslår vi en rak modell som direkt upptäcker ämnena i ett dokument genom att klustra sina ord , utan behov av en corpus. Vi definierar en vektorrepresentation som kallas normalized bag-of-topic-embeddings (nBTE) för att inkapsla de upptäckta ämnena och beräkna den mjuka cosinuslikheten mellan två nBTE-vektorer som dokumentets likhet. Dessutom föreslår vi en logistisk ordbetydelsefunktion som tilldelar ord olika viktvikter baserat på relativ diskriminerande kraft.Vår modell är effektiv när det gäller den genomsnittliga tidskomplexiteten. nBTE-representationen är också tolkbar som möjliggör ämnesidentifiering av dokumentet. På tremärkta offentliga dataset uppnådde vår modell jämförbar närmaste grannklassningsnoggrannhet med fem toppmoderna modeller. Vidare härledde vi från de tre dataseten fyra multi-ämnesdatasatser där varje etikett hänvisar till en uppsättning ämnen. Vår modell överensstämmer överens med de högteknologiska baslinjemodellerna med en stor marginal av fyra utmanande multi-ämnesdatasatser. Dessa arbetsstöd ger svar på forskningsproblemet av tisthesis:Kan vi konstruera en tolkbar dokumentrepresentation genom att klustra orden i ett dokument och effektivt och effektivt uppskatta dokumentets likhet?

No description available.

京都大学 / 新制・課程博士 / 博士(情報学) / 甲第24036号 / 情博第792号 / 新制||情||134(附属図書館) / 京都大学大学院情報学研究科社会情報学専攻 / (主査)准教授 馬 強, 教授 田島 敬史, 教授 森 信介 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM