Des milliers d'œuvres musicales sont décrites dans des catalogues des institutions culturelles, dont le rôle est de stocker toutes les créations musicales à travers le catalogage et de les diffuser auprès du grand public. Cette thèse s’inscrit dans le cadre du projet ANR DOREMUS -DOnnées en REutilisation pour la Musique en fonction des USages- qui vise à explorer les métadonnées des catalogues de trois grandes institutions culturelles : Bibliothèque Nationale de France (BNF), Philharmonie de Paris et Radio France afin qu'elles puissent communiquer entre elles et être mieux utilisées par les différents publics. Dans cette thèse, nous nous intéressons aux liens dits d’identité, exprimant une équivalence entre deux ressources différentes décrivant la même entité du monde réel. Notre objectif principal est de proposer une approche de liage générique, traitant certains challenges, avec comme cas concret d’utilisation les données de DOREMUS.Dans cette thèse, nous nous focalisons sur trois principaux challenges : (1) réduire la configuration manuelle de l’outil de liage, (2) faire face à différents types d’hétérogénéité entre les descriptions, et (3) Supprimer l’ambiguïté entre les ressources très similaires dans leur descriptions mais qui ne sont pas équivalentes. Certaines approches de liage demandent souvent l’intervention de l’utilisateur pour configurer certains paramètres. Ceci peut s’avérer être une tâche coûteuse pour l’utilisateur qui peut ne pas être expert du domaine. Par conséquent, une des questions de recherche que nous nous posons est comment réduire autant que possible l’intervention humaine dans le processus de liage des données. De plus, les descriptions des ressources peuvent présenter diverses hétérogénéités qu’un outil doit savoir gérer. Par ailleurs, les descriptions peuvent être exprimées dans différentes langues naturelles, avec des vocabulaires différents ou encore avec des valeurs différentes. La comparaison peut alors s’avérer très difficile en raison des variations selon trois dimensions : basées sur les valeurs, ontologiques et logiques. Dans cette thèse, nous analysons les aspects d’hétérogénéité les plus récurrents en identifiant un ensemble de techniques qui peuvent leur être appliquées. Un autre défi est la distinction entre des descriptions de ressources fortement similaires mais non équivalentes. En leur présence, la plupart des outils existants se voient diminuer leur efficacité en terme de qualité, en générant beaucoup de faux positifs. Dans cette optique, certaines approches ont été proposées pour identifier un ensemble de propriétés discriminatives appelées des clefs. De telles approches découvrent un très grand nombre de clés. La question qui se pose est de savoir si toutes les clés permettent de découvrir les mêmes paires d’instances équivalentes, ou si certaines sont plus significatives que d'autres. Aucune approche ne fournit de stratégie pour classer les clefs générées en fonction de leur efficacité à découvrir les bons liens. Afin d’assurer des alignements de qualité, nous avons proposé dans ce travail une nouvelle approche de liage de données visant à relever les défis décrits ci-dessus.Un outil de liage automatique de données hétérogènes, nommé Legato, qui répond aux challenges évoqués précédemment a été développé. Il est basé sur la notion de profile d’instance représentant chaque ressource comme un document textuel de littéraux gérant une variété d’hétérogénéités de données sans l’intervention de l’utilisateur. Legato implémente également une étape de filtrage de propriétés dites problématiques permettant de nettoyer les données du bruit susceptible de rendre la tâche de comparaison difficile. Pour pallier au problème de distinction entre les ressources similaires dans leur description, Legato implémente un algorithme basé sur la sélection et le ranking des clefs afin d’améliorer considérablement la précision au niveau des liens générés. / This thesis is part of the ANR DOREMUS project. We are interested in the catalogs of three cultural institutions: BNF (Bibliothèque Nationale de France), Philharmonie de Paris and Radio France, containing detailed descriptions about music works. These institutions have adopted the Semantic Web technologies with the aim of making these data accessible to all and linked.The links creation becomes particularly difficult considering the high heterogeneity between the descriptions of the same entity. In this thesis, our main objective is to propose a generic data linking approach, dealing with certain challenges, for a concrete application on DOREMUS data. We focus on three major challenges: (1) reducing the tool configuration effort, (2) coping with different kinds of data heterogeneities across datasets and (3) dealing with datasets containing blocks of highly similar instances. Some of the existing linking approaches often require the user intervention during the linking process to configure some parameters. This may be a costly task for theuser who may not be an expert in the domain. Therefore, one of the researchquestions that arises is how to reduce human intervention as much as possible inthe process of data linking. Moreover, the data can show various heterogeneitiesthat a linking tool has to deal with. The descriptions can be expressed in differentnatural languages, with different vocabularies or with different values. The comparison can be complicated due to the variations according to three dimensions: value-based, ontological and logical. Another challenge is the distinction between highly similar but not equivalent resource descriptions. In their presence, most of the existing tools are reduced in efficiency generating false positive matches. In this perspective, some approaches have been proposed to identify a set of discriminative properties called keys. Very often, such approaches discover a very large number of keys. The question that arises is whether all keys can discover the same pairs of equivalent instances, or ifsome are more meaningful than others. No approach provides a strategy to classify the keys generated according to their effectiveness to discover the correct links.We developed Legato — a generic tool for automatic heterogeneous data linking.It is based on instance profiling to represent each resource as a textual documentof literals dealing with a variety of data heterogeneities. It implementsa filtering step of so-called problematic properties allowing to clean the data ofthe noise likely to make the comparison task difficult. To address the problem ofsimilar but distinct resources, Legato implements a key ranking algorithm calledRANKey.

Estudamos uma classe de sistemas elípticos - \'elipson POT 2\' \'DELTA\' u + u = g(v) em \'ÔMEGA\' - \'elipson POT 2\' \'DELTA\' v + v f(u) em ÔMEGA \' PARTIAL\'u SOBRE \'PARTIAL n = \'PARTIAL v SOBRE PARTIAL n = O sobre \"PARTIAL\'\' ÔMEGA\' onde \' ÔMEGA ESTA CONTIDO EM R POT. N\' é um domínio limitado, com bordo regular e N \' > ou =\' 3. As não linearidades f e g são funções com crescimento superlinear e subcrítico no infinito. Estudamos resultados sobre a existência de uma sequência de soluções que se concentram, quando o parâmetro \'epsilon\' tende a zero, em um ponto da fronteira que maximiza a sua curvatura. Para isso utilizamos um resultado abstrato sobre existência de pontos críticos para funcionais fortemente indefinidos / We study an singularly perturbed Hamiltonean elliptic system - \'elipson POT 2\' \'DELTA\' u + u = g(v) in \'ÔMEGA\' - \'elipson POT 2\' \'DELTA\' v + v f(u) in ÔMEGA \' PARTIAL\'u ON \'PARTIAL n = \'PARTIAL v ON PARTIAL n\' = O sobre \"PARTIAL\'\' ÔMEGA\' when \'ÔMEGA THIS CONTAINED R POT. N\' is a smooth bounded domain, N \' > or =\' 3 and f and g are nonlinearities having superlinear and subcritical growth at infinity. We study an abstract result about existence of critical points of strongly as \' epsilon\' goes to zero, at a point of the boundary which maximizes the mean curvature of the boundary

La gestion des ressources humaines est une tâche importante pour toutes les organisations. Avec le nombre de candidatures en augmentation grâce à plusieurs plateformes en ligne, il est souhaitable de faire correspondre automatiquement les candidats avec des offres d’emploi. Les approches existantes utilisent les CVs sans compléter les informations par des recherches sur le Web, notamment le Web social. L’objectif de cette thèse est de surmonter cette limitation et proposer des méthodes pour découvrir des ressources en ligne pertinentes pour un demandeur d’emploi. À cet égard, une nouvelle méthode pour l’extraction d’informations clés à partir des CVs est proposée. Il s’agit d’un problème difficile puisque les CVs peuvent être multilingues et avoir des structures assez variées. En plus, les entités présentes sont suivant ambiguës. L’identification et la réconciliation des ressources en ligne en utilisant les informations clés sont un autre défi. Nous proposons un algorithme pour générer des requêtes et classer les résultats pour obtenir les ressources en ligne les plus pertinentes pour un demandeur d’emploi.. En outre, nous abordons spécifiquement la réconciliation de profils dans les réseaux sociaux grâce à une méthode qui est capable d’identifier les profils de individus à travers différents réseaux. Cette méthode utilise notamment les informations relatives à la localisation géographique des profils. A cet égard, nous proposons un algorithme permettant de désambiguïser les toponymes utilisés dans les profils pour indiquer une localité géographique ; cet algorithme peut être également utilisé pour inférer la localité d’un individu lorsqu’il ne l’a pas renseignée. Des expériences sur des ensembles de données réelles sont menées pour tous les différents algorithmes proposés dans cette thèse qui montrent de bons résultats. / Finding the appropriate individual to hire is a crucial part of any organization. With the number of applications increasing due to the introduction of online job portals, it is desired to automatically match applicants with job offers. Existing approaches that match applicants with job offers take resumes as they are and do not attempt to complete the information on a resume by looking for more information on the Internet. The objective of this thesis is to fill this gap by discovering online resources pertinent to an applicant. To this end, a novel method for extraction of key information from resumes is proposed. This is a challenging task since resumes can have diverse structures and formats, and the entities present within are ambiguous. Identification of Web results using the key information and their reconciliation is another challenge. We propose an algorithm to generate queries, and rank the results to obtain the most pertinent online resources. In addition, we specifically tackle reconciliation of social network profiles through a method that is able to identify profiles of individuals across different networks. Moreover, a method to resolve ambiguity in locations, or predict it when absent, is also presented. Experiments on real data sets are conducted for all the different algorithms proposed in this thesis and they show good results.

L'épuisement des ressources d'origine pétrochimique conduit à la recherche de matières premières renouvelables pour l'élaboration des membranes. La gélatine, un biopolymère abondant, sous produit de l'industrie agroalimentaire, représente grâce à ses propriétés filmogènes un candidat de choix pour l'élaboration de membranes “vertes”. L'objectif de cette thèse est l'élaboration de membranes à base de gélatine et l'étude de l'impact de la structure du matériau sur les propriétés mécaniques, thermiques, de résistance à l'eau et de transfert gazeux. Pour cela les mécanismes d'élaboration par TIG-Dry cast process ont d'abord été formalisés par l'établissement du diagramme de phase du système gélatine/eau. Puis des réticulants alternatifs au glutaraldéhyde, toxique, ont été examinés pour augmenter la résistance à l'eau de la gélatine hydrosoluble. L'acide férulique et le téréphthalaldéhyde se sont montrés les plus prometteurs et complémentaires suivant les applications visées. Les films de gélatine se sont révélés barrières aux gaz à cause de la forte cristallinité induite par la renaturation des triples hélices. L'ajout d'un élastomère de la famille des polyéther amines a permis d'augmenter drastiquement les coefficients de perméabilité du CO2 de 1,4 à 250 Barrer. L'influence de la température et de l'humidité relative des flux gazeux sur les perméabilités a également été étudiée. / Petroleum based raw materials shortage leads to investigate renewable raw materials for membrane elaboration. Gelatin, an abundant, industrial by-product is a biosourced polymer with filmogenic properties which makes it an educated choice for “green” membrane production. This thesis work aims at developing gelatin based membrane and studying the influence of the material structure on mechanical and thermal properties, water resistance and gas transport properties. Thus, the elaboration mechanisms by TIG/Dry-cast process were studied in details by establishing the phase diagram of the gelatin/water system. To improve the water resistance of the hydrosoluble gelatin, crosslinking is necessary. Alternative cross-linkers were tested to replace the glutaraldehyde, classified as toxic. Ferulic acid and terephthalaldehyde were promising and showed complementary characteristics. The high crystallinity level of gelatin films, related to their renaturation level, led to rather gas barrier properties. By adding an elastomer, polyetheramine, the permeability to CO2 increased from 1.4 to an outstanding 250 Barrer. The influence of the temperature and relative humidity of the gas flux on permeability was also studied.

Dans une configuration optimale, la couche photo-active d’une cellule solaire organique forme une hétérojonction volumique entre un matériau donneur et un matériau accepteur d’électrons. Cette morphologie optimale se trouve dans un état thermodynamique métastable. En cours de fonctionnement, l’absorption incidente provoque une élévation de la tempéra- ture des dispositifs. L’apport d’énergie thermique au système le fait évoluer vers un état thermodynamique stable correspondant à une micro-séparation de phase entre les deux matériaux nuisible aux performances photovoltaïques du dispositif. La solution à cette problématique envisagée dans cette thèse est de figer la morphologie optimale de la couche photo-active par réticulation chimique. Dans le but de prévenir la diffusion et la cristallisation des dérivés du fullerène, plusieurs approches ont été étudiées : la formation d’une maille de polymères réticulés, l’accroche des dérivés du fullerène aux chaînes du polymères ou la liaison entre plusieurs dérivés du fullerène. Les différentes fonctions réticulantes utilisées dans ces approches sont l’allyle, le cinnamate, l’anthracène et l’azoture. Au cours de ces travaux, des résultats satisfaisants ont été obtenus avec la fonction azoture. Basé sur cette fonction, nous avons développé un additif qui permet d’augmenter notablement la stabilité thermique des cellules solaires réticulées. Cette approche a été validée sur plusieurs couples de matériaux polymères/dérivés du fullerène à l’état de l’art. / Efficient bulk-heterojunction solar cells require an interpenetrating network of electron donating and electron accepting materials. Indeed, the optimum nano-sized phase segregation is a thermodynamically metastable system. Under operating conditions, especially under solar illumination, the temperature of the solar cell increases significantly. Adding thermal energy to the system leads to further micro-phase separation that is harmful for photovoltaic device performances. In order to freeze the optimal morphology, several approaches based on chemical cross-linking have been studied : formation of cross-linked polymer network, binding fullerene derivatives to polymer side chains or cross-linking between fullerene derivatives. Different cross-linkable functions have been used : allyl, cinamate, anthracene and azide. A versatile stabilization strategy of bulk heterojunctions morphology has been found by mean of incorporating a cross-linkable additive in low amounts with different polymer :fullerene blends. This approach is validated on multiple materials at the state-of-the-art.

The thesis analyses together used instruments of marketing communication mix of town Pelhřimov. These instruments are evaluated from the view of informing and linking of citizens into the decision-making process in the public service. By Questionnaire investigation the thesis recognizes the effectivity of marketing instruments and their impact on target groups of inhabitants of the town. It also identifies possible communication barriers and gathers suggestions for new forms of communication. On the basis of the synthesis of found facts it puts before measures for improving of current status of marketing communication of town Pelhřimov.

Chemical biology can be defined as the area of science where chemical tools are used to study biological systems. The simplest way this can be achieved is in the identification of compounds which inhibit or modulate a biological pathway and the consequences studied. However, novel tools are required to enable, for example, the development of assays to allow simpler screening of difficult targets such as membrane proteins and protein-protein interactions. A series of kisspeptin analogues were synthesised for the development of a screening platform compatible with G-protein coupled receptors and tagged one bead one compound (OBOC) combinatorial libraries. Fluorescently labelled kisspeptin showed good affinity for GPCR54 and an on-bead version of the peptide, with the required C-terminal amide presented away from the bead was prepared and used for testing possible screening methods. GPCR54 was expressed in a number of formats and a kisspeptin based OBOC library designed and synthesised. Investigation into the C-terminal RF-amide motif of Kisspeptin was also carried out in order to assess the importance of the carbonyl moiety. The corresponding peptide amine was synthesised and the compound biologically assessed. This led to the development of a novel acid labile benzofuranone (ALBA) linker for anchoring amines to a solid support. For the preparation of fluorescent kisspeptin ligands, a novel general synthetic route which gives direct access to single isomer functionalised rhodamine dyes from phthalides has been developed. This circumvents the arduous task of isomer separation usually associated with the synthesis of functionalised rhodamines. The route has been demonstrated with a range of linkage groups and rhodamine types. This rhodamine material was used as a reporter group in various multifunctional reagents synthesised using a trifunctional orthogonally protected backbone (TOBa), which was prepared on a solid support and enables rapid synthesis of trifunctional reagents. This resin takes advantage of protecting group orthogonality and the high yields of peptide bond formation. A series of trifunctional reagents for screening use were prepared using this resin. A proof of concept study was carried out involving the simultaneous labelling and immobilisation of a protein for applications in probing protein-protein interactions. Development of a trifunctional hydroxamic acid containing cross-linker was carried out which takes advantage of its reaction with boronic acids to enable reversible capture on solid support for enrichment of cross-linked peptides. A new benzophenone based heterobifunctional reagent was prepared for protein cross-linking and mass spectrometry analysis. This was shown to give complimentary reactivity to existing cross-linkers, allowing more structural information to be extracted from protein samples.

3D proteomics combines chemical cross-linking with mass spectrometry to study the structure of protein assemblies and protein-protein interactions both in vitro and in vivo by providing distance constraints that indicate which residues are in close spatial proximity. I addressed the main bottleneck of this technology: the reliable identification of cross-linked peptides. Reporter ion signatures for cross-linked peptides were developed, by fragmenting model compounds containing two lysine residues joined by a cross-linker backbone or a lysine residue modified with a hydrolysed cross-linker. The reporter ion signatures showed 97% specificity at 90% sensitivity and segregated cross-linked from modified and linear peptides. They decreased the false discovery rate of the identification of cross-linked peptides from 5% to 1% in a large dataset. The signatures permit data sorting during and after mass spectrometry acquisition. The advanced 3D proteomics workflow was applied to study the protein-protein interactions in Mycoplasma pneumoniae cells. In lysates of the bacterium we identified 128 protein-protein interactions (of which 24 are novel) and obtained in vivo topological data on 208 proteins, even for cases where high-resolution structures are not yet available. We showed that our data are in excellent agreement with crystal structures of proteins and complexes where available. We defined a network of ribosomal and RNA polymerase proteins that reveals an intricate link between transcription and translation in bacteria. We demonstrated that the method is suitable for identification of homomultimeric protein complexes by exploiting peptide pairs of identical amino acid sequence. The technology has the potential to provide a complete protein interaction network map after the selective enrichment of cross-lined peptides is achieved. The method was next applied to investigate the structure of condensin and cohesin complexes, which play a crucial role in stabilization of chromosome structure during mitosis. The complexes were purified, cross-linked and their linkage map created. The condensin coiled coil cross-linked on the entire length was modeled. The information was used to direct the analysis of in situ cross-linked condensin in intact chromosomes. I found two high confidence linkages between SMC2 and SMC4 coiled coils and identified H2A as a potential condensin receptor on chromosomes.

<p>Predmet doktorske disertacije je izučavanje različitih bioprocesa za modifikovanje biljnih globulina radi unapređenja njihovih funkcionalnih karakteristika. Istraživanja su zasnovana na karakterizaciji i enzimskoj modifikaciji glavnog rezervnog proteina (12S), kukurbitina, iz semena uljane tikve (<em>Cucurbita pepo</em>). Osnova istraživanja je enzimska konverzija globulina i dobijanje proteinskih modifikata delovanjem hidrolaza i transferaza. U okviru istraživanja, enzimski procesi modifikacije globulina izučavani su sa dva aspekta: enzimska hidroliza i enzimsko umrežavanje (cross-linking), primenom komercijalnih enzimskih preparata. Takođe istraživanja obuhvataju i razvoj i kontrolu samih bioprocesa definisanjem i optimizacijom procesnih parametara (temperature, pH, koncentracije enzima i supstrata, vreme reakcije). Ovako definisani procesi eksploatisani su u cilju kreiranja željenih funkcionalnih karakteristika proteina spram njihove potencijalne primene u formulacijama hrane. Odabir i optimizacija procesnih parametara i modelovanje bioprocesa izvedeno je implementiranjem nove kompjuterske i analitičke metodologije</p> / <p>The PhD thesis research is aimed at development of different bioprocesses for modification of plant globulins in order to improve their functional properties. Studies are based on characterization and enzymatic modification of major storage protein (12S), cucurbitin derived from pumpkin oil seed (<em>Cucurbita pepo</em>). The base of research is enzymatic conversion of cucurbitin by hydrolase and transferase. Two different enzymatic processes are used for protein modification: (i) enzymatic hydrolysis and (ii) enzymatic cross-linking. To monitor, control the bioprocesses, and definition of process parameters, such as temperature, pH, enzyme-substrate ratio, reaction time, Response Surface Methodology (RSM) was used. In addition, RSM was employed for production of protein modification with desired functional properties.</p>

All current theories of spoken word recognition (e.g., Allopenna et al., 1998; McClelland & Elman, 1986; Norris, 1994) suggest that any part of a target word triggers activation of candidate words. Visual world paradigm studies have relied on the linking hypothesis that the probability of looking at the referent of a word directly tracks the word’s level of activation (e.g., Allopenna et al., 1998). However, how much information is needed to trigger a saccade to a visual representation of the word’s referent? To address this question, the present study manipulated the number and location of shared segments between the target and competitor words. Experimental evidence is provided by two visual world paradigm experiments on Japanese, using natural and synthesized speech. In both experiments, cohort competitor pictures were not fixated more than unrelated distractor pictures unless the cohort competitor shares the initial CVC with the target. Bayesian analyses provide strong support for the null hypothesis that shorter overlap does not affect eye movements. The results suggest that a listener needs to accumulate enough evidence for a word before a saccade is generated. The human data were validated by an interactive computational model (TRACE: McClelland & Elman, 1986). The model was adapted to Japanese language to examine whether the TRACE model predicts competitor effects that fit human data. The model predicted that there should be effects when words share any amount with a target which confirms the current theory. However, the model did not fit the human data unless there is longer overlap between words. This indicates that eye movements are not as closely tied to fixation probabilities of lexical representations as previously believed. The present study suggests that looking at a referent of a word is a decision, made when the word’s activation exceeds a context-specific threshold. Subthreshold activations do not drive saccades. The present study conclude that decision-making processes need to be incorporated in models linking word activation to eye movements.