Content-based image indexing and retrieval (CBIR) system needs to consider several types of visual features and spatial information among them (i.e., different point of views) for better image representation. This thesis presents a novel approach that exploits an extension of the language modeling approach from information retrieval to the problem of graph-based image retrieval. Such versatile graph model is needed to represent the multiple points of views of images. This graph-based framework is composed of three main stages: Image processing stage aims at extracting image regions from the image. It also consists of computing the numerical feature vectors associated with image regions. Graph modeling stage consists of two main steps. First, extracted image regions that are visually similar will be grouped into clusters using an unsupervised learning algorithm. Each cluster is then associated with a visual concept. The second step generates the spatial relations between the visual concepts. Each image is represented by a visual graph captured from a set of visual concepts and a set of spatial relations among them. Graph retrieval stage is to retrieve images relevant to a new image query. Query graphs are generated following the graph modeling stage. Inspired by the language model for text retrieval, we extend this framework for matching the query graph with the document graphs from the database. Images are then ranked based on the relevance values of the corresponding image graphs. Two instances of the visual graph model have been applied to the problem of scene recognition and robot localization. We performed the experiments on two image collections: one contained 3,849 touristic images and another composed of 3,633 images captured by a mobile robot. The achieved results show that using visual graph model outperforms the standard language model and the Support Vector Machine method by more than 10% in accuracy.

Cette thèse propose un système d'aide à l'indexation et à la recherche de documents pédagogiques fondé sur l'utilisation de Wikipédia.l'outil d'aide à l'indexation permet de seconder les documentalistes dans la validation, le filtrage et la sélection des thématiques, des concepts et des mots-clés issus de l'extraction automatique d'un document. En effectuant une analyse des données textuelles d'un document, nous proposons au documentaliste une liste de descripteurs permettant de représenter et discriminer le document. Le travail du documentaliste se limite alors à une lecture rapide du document et à la sélection et suppression des descripteurs suggérés par le système pour rendre l'indexation homogène, discriminante et exhaustive. Pour cela nous utilisons Wikipédia comme base de connaissances. Le modèle utilisé pour l'extraction des descripteurs permet également de faire de la recherche d'information sur un corpus de document déjà indexé.

Thesis (Ph. D.)--University of California, San Diego, 2007. / Title from first page of PDF file (viewed October 3, 2007). Available via ProQuest Digital Dissertations. Vita. Includes bibliographical references (p. 60-64).

Investors have historically demanded a return premium for taking on the risk of illiquidity both in terms of characteristic and systematic liquidity risk. Recent research have presented results suggesting that the liquidity premium is diminishing. The increasing popularity of passive investments such as Exchange Traded Funds (ETFs) have been proposed as a driving force for the declining trend. Despite the popularity of ETFs, there is limited research how they impact the financial markets. The purpose of this thesis is to investigate how the liquidity premium has developed in the United States between 1997 and 2016 and to explore if developments in the liquidity premium can be linked to the capital inflow to the United States ETF market. The thesis uses measures of stocks’ spreads and order book depths as proxies for the characteristic and systematic liquidities. The proxies are used to test if liquidity has influenced stock returns over 1-year, 5-years and the entire 20-year period. The empirical results obtained through Fama-MacBeth regressions show that the liquidity premium can fluctuate by both sign and magnitude year by year. The characteristic risk premium is negative and significant for the entire 20-year period and the 1-year regressions suggests a clear negative trend. The systematic liquidity premium on the other hand is positive and significant for the entire 20-year period but the 1-year regressions do not show a clear trend. The empirical results show no statistical significance that ETFs influence the liquidity premium. However, the graphical interpretation of the 1-year regressions suggests that the characteristic liquidity premium is negatively correlated with the growth of ETFs. The negative characteristic premium implies that investors are not being adequately compensated for the risk of illiquidity and should therefore avoid a liquidity-based investing strategy which has generated excess return in the past.

L’objectif principal de cette thèse est de montrer que les informations lexicales issues d’un dictionnaire de langue, tel le Trésor de la langue française informatisé (TLFi), peuvent améliorer les processus d’indexation et de recherche d’images. Le problème d’utilisation d’une telle ressource est qu’elle n’est pas suffisamment formalisée pour être exploitée d’emblée dans un tel domaine d’application. Pour résoudre ce problème, nous proposons, dans un premier temps, une approche de construction automatique de hiérarchies sémantiques à partir du TLFi. Après avoir défini une caractéristique quantitative (mesurable) et comparable des noms apparaissant dans les définitions lexicographiques, à travers une formule de pondération permettant de sélectionner le nom de poids maximal comme un bon candidat hyperonyme pour un lexème donné du TLFi, nous proposons un algorithme de construction automatique de hiérarchies sémantiques pour les lexèmes des vocables du TLFi. Une fois notre approche validée à travers des évaluations manuelles, nous montrons, dans un second temps, que les hiérarchies sémantiques obtenues à partir du TLFi peuvent être utilisées pour l’enrichissement d’un thésaurus construit manuellement ainsi que pour l’indexation automatique d’images à partir de leurs descriptions textuelles associées. Nous prouvons aussi que l’exploitation d’une telle ressource dans le domaine de recherche d’images améliore la précision de la recherche en structurant les résultats selon les domaines auxquels les concepts de la requête de recherche peuvent faire référence. La mise en place d’un prototype nous a permis ainsi d’évaluer et de valider les approches proposées. / The main purpose of this thesis is to show that the lexical information issuing from a language dictionary, as the Trésor de la langue française informatisé (TLFi), can improve the image indexing and retrieval process. The problem of using of such resource is that it is not sufficiently formalized to be exploited immediately in such application domain.To solve this problem, we propose a first approach of automatic construction of semantic hierarchies from TLFi. After defining a quantitative (measurable) and comparable characteristic of names appearing in dictionary definitions, through a weighting formula that allows us to select the name of the maximum weight as a good hypernym candidate for a given TLFi lexeme, we suggest an algorithm of automatic construction of semantic hierarchies for the lexemes of TLFi vocables.Once our approach is validated through manual evaluations, we demonstrate in the second time that the semantic hierarchies obtained from TLFi can be used to enrich a thesaurus manually built as well as for automatic image indexing using their associated text descriptions. We also prove that the use of such resource in the domain of image retrieval improves the accuracy of search by structuring the results according the domains to which the concepts of the search query are related to. The implementation of a prototype allowed us to evaluate and validate the proposed approaches.

Hakhun Tangsa is one of around eighty ethnic and linguistic communities called Tangsa or Tangshang. Hakhuns live mostly in Arunachal Pradesh, India, and in Sagaing Division, Myanmar. The number of speakers is estimated at around ten thousand. Hakhun is a Tibeto-Burman language, and it forms a subgroup with Nocte, Wancho, Phom, Konyak, Chang, and Khiamngan called Konyak or Northern Naga. Hakhun is a tonal language with twenty-two consonants, six vowels, and a simple syllable structure. Open word classes include Nouns and Verbs; property concept terms form a subclass of verbs. Noun roots are mostly monosyllabic, and most multisyllabic nouns are compounds. Nominal morphology includes prossessive prefixes and a set of semantically specific suffixes. Case is coded by postpositions. Verb roots are also mostly monosyllabic. A few verbs have suppletive stems. Verb serialization is common, and expresses complex events like resultative and sequential. A few grammaticalized verbs/elements contribute abstract meanings like phase, associated motion, causative, benefactive, etc. Typical verbal categories are expressed by independent particles. The most extensive and grammatically obligatory set consists of single syllable particles called operators, which express verbal categories like tense, mood, deixis, negation, inverse, and argument indexation. The typical argument indexation pattern is hierarchical. Deviations from this pattern is used to express certain pragmatic effects like affectedness and politeness. Non-verbal clauses may take overt copulas depending on tense and polarity. Most semantic distinctions, such as equation, property-concepts, quantification, simulation, and location are expressed by the nominal strategy. Existential and possession are expressed by a distinct strategy. Typical verbal clauses include intransitive, transitive, and ditransitive; less typical ones include weather condition, sensation-emotion, reflexive, reciprocal, and ‘need’ constructions. Person-based split-ergativity is seen in case marking, where first and second person singular arguments follow accusative, and the rest ergative alignment. Accusative alignment is also found in argument indexation in non-final clauses. The object alignment is indirective in case marking. Complement clauses include sentence-like, non-finite, and infinitive complement clauses. Adverbial clauses include various kinds of temporal clauses, temporal/conditional clauses, counterfactual, concessive, purpose, and substitutive clauses. Clause chaining (medial-final) is prevalent. Independent sentences are linked through tail-head linking and through connectives.

Avec l’avènement de « l'ère des données », les besoins des systèmes de traitement de l'information en ressources de calcul ont explosé, dépassant largement les évolutions technologiques des processeurs modernes. Dans le domaine de l'apprentissage statistique en particulier, les paradigmes de calcul massivement distribués représentent la seule alternative praticable.L'algorithmique distribuée emprunte la plupart de ses concepts à l'algorithmique classique, centralisée et séquentielle, dans laquelle le comportement du système est décrit comme une suite d'instructions exécutées l'une après l'autre. L'importance de la communication entre unités de calcul y est généralement négligée et reléguée aux détails d'implémentation. Or, lorsque le nombre d'unités impliquées augmente, le poids des opérations locales s'efface devant les effets émergents propres aux larges réseaux d'unités. Pour conserver les propriétés désirables de stabilité, de prédictibilité et de programmabilité offertes par l'algorithmique centralisée, les paradigmes de calcul distribué doivent dès lors intégrer cette dimension qui relève de la théorie des graphes.Cette thèse propose un cadre algorithmique pour l'apprentissage statistique large échelle, qui prévient deux défaut majeurs des méthodes classiques : la centralisation et la synchronisation. Nous présentons ainsi plusieurs algorithmes basés sur des protocoles Gossip décentralisés et asynchrones, applicables aux problèmes de catégorisation, estimation de densité, réduction de dimension, classification et optimisation convexe. Ces algorithmes produisent des solutions identiques à leurs homologues centralisés, tout en offrant une accélération appréciable sur de larges réseaux pour un coût de communication très réduit. Ces qualités pratiques sont démontrées mathématiquement par une analyse de convergence détaillée. Nous illustrons finalement la pertinence des méthodes proposées sur des tâches d'indexation multimédia et de classification d'images. / With the advent of the "data era", the amount of computational resources required by information processing systems has exploded, largely exceeding the technological evolutions of modern processors. Specifically, contemporary machine learning applications necessarily resort to massively distributed computation.Distributed algorithmics borrows most of its concepts from classical centralized and sequential algorithmics, where the system's behavior is defined as a sequence of instructions, executed one after the other. The importance of communication between computation units is generally neglected and pushed back to implementation details. Yet, as the number of units grows, the impact of local operations vanishes behind the emergent effects related to the large network of units. To preserve the desirable properties of centralized algorithmics such as stability, predictability and programmability, distributed computational paradigms must encompass this graph-theoretical dimension.This thesis proposes an algorithmic framework for large scale machine learning, which prevent two major drawbacks of classical methods, namely emph{centralization} and emph{synchronization}. We therefore introduce several new algorithms based on decentralized and asynchronous Gossip protocols, for solving clustering, density estimation, dimension reduction, classification and general convex optimization problems, while offering an appreciable speed-up on large networks with a very low communication cost. These practical advantages are mathematically supported by a theoretical convergence analysis. We finally illustrate the relevance of proposed methods on multimedia indexing applications and real image classification tasks.

L'objectif de cette thèse est de proposer une solution d'indexation ``scalable'' et basée objet de flux vidéos HD compressés avec Motion JPEG2000. Dans ce contexte, d'une part, nous travaillons dans le domaine transformé hiérachique des ondelettes 9/7 de Daubechies et, d'autre part, la représentation ``scalable'' nécessite des méthodes en multirésolution, de basse résolution vers haute résolution. La première partie de ce manuscrit est dédiée à la définition d'une méthode d'extraction automatique des objets en mouvement. Elle repose sur la combinaison d'une estimation du mouvement global robuste et d'une segmentation morphologique couleur à basse résolution. Le résultat est ensuite affiné en suivant l'ordre des données dans le flux scalable. La deuxième partie est consacrée à la définition d'un descripteur sur les objets précédemment extraits, basé sur les histogrammes en multirésolution des coefficients d'ondelettes. Enfin, les performances de la méthode d'indexation proposée sont évaluées dans le contexte de requêtes scalables de recherche de vidéos par le contenu. / This thesis aims at proposing a solution of scalable object-based indexing of HD video flow compressed by MJPEG2000. In this context, on the one hand, we work in the hierarchical transform domain of the 9/7 Daubechies' wavelets and, on the other hand, the scalable representation implies to search for multiscale methods, from low to high resolution. The first part of this manuscript is dedicated to the definition of a method for automatic extraction of objects having their own motion. It is based on a combination of a robust global motion estimation with a morphological color segmentation at low resolution. The obtained result is then refined following the data order of the scalable flow. The second part is the definition of an object descriptor which is based on the multiscale histograms of the wavelet coefficients. Finally, the performances of the proposed method are evaluated in the context of scalable content-based queries.

In recent decades, many investors have abandoned hopes of achieving above market returns through active management, and consigned themselves to passive investing in the form of market capitalization based portfolios. Using Swedish stock exchange data from 2002-2016, this thesis investigates if there is a way to harmonize the strengths of active management, yielding potential above market returns, and passive index investing, implying lower fees and transparency. Based on observations from 275 companies, analysed through market model regressions, the results suggest that fundamentally invested value and quality portfolios create an alpha of 1-2 percent quarterly relative the market capitalization benchmark portfolio. Moreover, the results constitute basis for performing real investments, as they take into consideration the transaction costs implied by portfolio turnover. Furthermore, the findings of greater risk-adjusted returns through fundamentally weighted portfolios stand in opposition to the efficient market hypothesis.

Coordination of social security in the EU - indexation of family benefits in the Republic of Austria As a unique system of cooperation across the European territory, social security coordination is currently facing political and legal challenges, which culminated in the ongoing legal proceedings initiated by the European Commission as a plaintiff against one of the Member States, namely the Republic of Austria. The first part of this master thesis (chapters one and two) introduces the reader to the general characteristics of coordination, its principles and its anchoring in the EU law. Part two (chapters three to five) allows the reader to learn more about the issue of indexation of (coordinated) family benefits and its history in Europe. It looks in more detail at indexation in the Republic of Austria, specifically its journey from a political idea to an application practice that conflicts with the current EU coordination arrangements, in particular the principle of equal treatment. The first chapter deals with the historical development of the idea of social security coordination, an explanation of the concept, as well as the sources of EU and international law, their scope in this field and their interrelations. The chapter concludes by tracing the link between coordination and one of the EU's...