Dissertations / Theses on the topic 'Classement des documents texte'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Classement des documents texte.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Poulain, d'Andecy Vincent. "Système à connaissance incrémentale pour la compréhension de document et la détection de fraude." Thesis, La Rochelle, 2021. http://www.theses.fr/2021LAROS025.
Full textThe Document Understanding is the Artificial Intelligence ability for machines to Read documents. In a global vision, it aims the understanding of the document function, the document class, and in a more local vision, it aims the understanding of some specific details like entities. The scientific challenge is to recognize more than 90% of the data. While the industrial challenge requires this performance with the least human effort to train the machine. This thesis defends that Incremental Learning methods can cope with both challenges. The proposals enable an efficient iterative training with very few document samples. For the classification task, we demonstrate (1) the continue learning of textual descriptors, (2) the benefit of the discourse sequence, (3) the benefit of integrating a Souvenir of few samples in the knowledge model. For the data extraction task, we demonstrate an iterative structural model, based on a star-graph representation, which is enhanced by the embedding of few a priori knowledges. Aware about economic and societal impacts because the document fraud, this thesis deals with this issue too. Our modest contribution is only to study the different fraud categories to open further research. This research work has been done in a non-classic framework, in conjunction of industrial activities for Yooz and collaborative research projects like the FEDER Securdoc project supported by la région Nouvelle Aquitaine, and the Labcom IDEAS supported by the ANR
Sayadi, Karim. "Classification du texte numérique et numérisé. Approche fondée sur les algorithmes d'apprentissage automatique." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066079/document.
Full textDifferent disciplines in the humanities, such as philology or palaeography, face complex and time-consuming tasks whenever it comes to examining the data sources. The introduction of computational approaches in humanities makes it possible to address issues such as semantic analysis and systematic archiving. The conceptual models developed are based on algorithms that are later hard coded in order to automate these tedious tasks. In the first part of the thesis we propose a novel method to build a semantic space based on topics modeling. In the second part and in order to classify historical documents according to their script. We propose a novel representation learning method based on stacking convolutional auto-encoder. The goal is to automatically learn plot representations of the script or the written language
Limame, Mohamed. "Un Modèle généralisé pour le classement de documents multimédia." Grenoble 2 : ANRT, 1988. http://catalogue.bnf.fr/ark:/12148/cb376153269.
Full textLimame, Mohamed. "Un modele generalise pour le classement de documents multimedia." Paris 7, 1988. http://www.theses.fr/1988PA077192.
Full textDenoyer, Ludovic. "Apprentissage et inférence statistique dans les bases de documents structurés : application aux corpus de documents textuels." Paris 6, 2004. http://www.theses.fr/2004PA066087.
Full textDelecraz, Sébastien. "Approches jointes texte/image pour la compréhension multimodale de documents." Thesis, Aix-Marseille, 2018. http://www.theses.fr/2018AIXM0634/document.
Full textThe human faculties of understanding are essentially multimodal. To understand the world around them, human beings fuse the information coming from all of their sensory receptors. Most of the documents used in automatic information processing contain multimodal information, for example text and image in textual documents or image and sound in video documents, however the processings used are most often monomodal. The aim of this thesis is to propose joint processes applying mainly to text and image for the processing of multimodal documents through two studies: one on multimodal fusion for the speaker role recognition in television broadcasts, the other on the complementarity of modalities for a task of linguistic analysis on corpora of images with captions. In the first part of this study, we interested in audiovisual documents analysis from news television channels. We propose an approach that uses in particular deep neural networks for representation and fusion of modalities. In the second part of this thesis, we are interested in approaches allowing to use several sources of multimodal information for a monomodal task of natural language processing in order to study their complementarity. We propose a complete system of correction of prepositional attachments using visual information, trained on a multimodal corpus of images with captions
Belkhiter, Nadir. "Transducteur d'affichage de documents techniques décrits en LRDM(texte, graphique, fac-similé)." Lyon 1, 1985. http://www.theses.fr/1985LYO19006.
Full textWehbe, Hassan. "Synchronisation automatique d'un contenu audiovisuel avec un texte qui le décrit." Thesis, Toulouse 3, 2016. http://www.theses.fr/2016TOU30104/document.
Full textWe address the problem of automatic synchronization of an audiovisual content with a procedural text that describes it. The strategy consists in extracting pieces of information about the structure from both contents, and in matching them depending on their types. We propose two video analysis tools that respectively extract: * Limits of events of interest using an approach inspired by dictionary quantization. * Segments that enclose a repeated action based on the YIN frequency analysis method. We then propose a synchronization system that merges results coming from these tools in order to establish links between textual instructions and the corresponding video segments. To do so, a "Confidence Matrix" is built and recursively processed in order to identify these links in respect with their reliability
Raliarivony, Fara Lancha Catherine. "Elaboration d'un plan de cotation des collections en libre accès de la bibliothèque universitaire centrale de Lille 3 (Lettres, arts et sciences humaines)." [S.l.] : [s.n.], 2004. http://www.enssib.fr/bibliotheque/documents/ppp/pppraliarivony.pdf.
Full textRousseau, Emilie. "Etude d'une chaîne de publication Web de texte intégral au format XML." [S.l.] : [s.n.], 2001. http://www.enssib.fr/bibliotheque/documents/dessid/rsrousseau.pdf.
Full textSalah, Soraya Ollès Christian. "Mise en ligne en texte intégral des thèses de l'Université Paul-Sabatier." [S.l.] : [s.n.], 2004. http://www.enssib.fr/bibliotheque/documents/ppp/pppsalah.pdf.
Full textClavier, Annie. "La sémiologie de l’image et du texte dans les documents utilisés en pédagogie espagnole." Paris 10, 2010. http://www.theses.fr/2010PA100225.
Full textImages play a very significant part in today’s civilization. The iconic dimension is one of the fundamental aspects of a new culture in which pictures proliferate in everyday life – whether in books, papers, posters or on screens – and where they hold an important role along with texts, even in primary and secondary school handbooks and in children’s and teenagers’ fiction works. Pictures and writing share the same origin : they were merged on Paleolithic walls. Then, they have evolved separately and in parallel. Thanks to the current means of production and circulation of texts and images, and to 20th century discoveries in cognitive sciences, they are now merged again; they are both considered as means of communication. The studies in semiology have proved that pictures, just like language, are a system of signs which express ideas, produce meaning and which we use in order to communicate. Images therefore show numerous analogies with texts regarding their production, functioning and reception. But some differences between texts and images do not enable the pictures to replace the texts as the primary vehicles of communication and information because they need the texts to compensate some of their expressive lacks. Nowadays in Spain, as in France, the new pedagogical directions stress the command of oral and written languages in primary and secondary education. But they also acknowledge the pervasiveness and power of images in contemporary writing. Although their attractive, persuasive and, according to some people, negative characters are recognized, images are widely used in Spanish education along with the texts, in didactic works and in class activities alike. In education, images can be usefully employed along with the texts since they both have some specificities. Education sciences perceive the great efficiency of images used along with texts: by using the qualities and the lacks of each code – verbal and iconic – the two interact, they clarify and reinforce each other so as to better transmit knowledge to the pupils. As a democratic state, today’s Spain gives great importance to the education received by the young to become citizens. Spanish teachers take into account the surrounding culture and also use documents from outside the school which include all types of texts and images – such as newspapers, magazines, posters, websites – to teach the pupils how to act responsibly and have a critical mind regarding the contents, tools and media of their everyday environment. In the current Spanish pedagogical system, the teachers emphasize the “defense and protection” against the “manipulation” and apathy of the critical mind. But it has not always been so in Spain, as during the Franco era
Delacroix, Quentin. "Un système pour la recherche plein texte et la consultation hypertexte de documents techniques." Clermont-Ferrand 2, 1999. http://www.theses.fr/1999CLF2A001.
Full textThomas, Corinne. "Accès par le contenu à des documents numérisés contenant du texte et de l'image." Paris 7, 2001. http://www.theses.fr/2001PA077150.
Full textAyache, Stéphane. "Indexation de documents vidéos par concepts par fusion de caractéristiques audio, vidéo et texte." Grenoble INPG, 2007. http://www.theses.fr/2007INPG0071.
Full textWork deals with information retrieval and aims to reach semantic indexing of multimediaIments. The state of the art approach tackle this problem by bridging of the semantic gap between level features, from each modality, and high-Ievel features (concepts), which are useful for humans. We propose an indexing model based on networks of operators into which data flows, called numcepts, unify informations from the various modalities and extracted at several level of abstraction. We present an instance of this model where we describe a topology of the operators and the numcepts we have deveIoped. We have conducted experiments on TREC VIDEO corpora in order to evaluate various organizations of the networks and the choice of the operators. We have studied those effects on performance of concept detection. We show that a network have to be designed with respect to the concepts in order to optimize the indexing performance
Kowarski, Irène Bolliet Louis. "Les bases de données textuelles étude du concept de document et application à deux réalisations /." S.l. : Université Grenoble 1, 2008. http://tel.archives-ouvertes.fr/tel-00308638.
Full textOuji, Asma. "Segmentation et classification dans les images de documents numérisés." Phd thesis, INSA de Lyon, 2012. http://tel.archives-ouvertes.fr/tel-00749933.
Full textPhan, Huy-Khanh. "Contribution à l'informatique multilingue : extension d'un éditeur de documents structurés." Lille 1, 1991. http://www.theses.fr/1991LIL10044.
Full textROBERT, Laurent. "Annotation et visualisation interactives de documents hypermédias." Phd thesis, Télécom ParisTech, 2001. http://tel.archives-ouvertes.fr/tel-00005631.
Full textA cette problématique de l'utilisation des documents en milieu informatique, nous proposons deux approches complémentaires issues des domaines de l'interaction homme-machine et de la visualisation d'information.
La première approche repose sur le concept de lecture active qui vise à transformer le " lecteur " en " acteur " en lui permettant d'enrichir interactivement les documents qu'il est en train de lire par un ensemble de signes graphiques (surlignages, cerclages, etc.) et d'informations textuelles (notes, commentaires, etc.). L'idée consiste à permettre aux utilisateurs de construire leur propre système de repérage afin de faciliter la compréhension et la réutilisation des documents. Une attention particulière est portée à la manière d'associer ces méta-données " ajoutées " (que nous nommons, d'une manière générique, " annotations ") aux documents sources au moyen de liens hypertextes, ainsi qu'à leur codage informatique dans un format normalisé (dérivé de XML).
La seconde approche consiste à faciliter l'exploration d'espaces documentaires en exploitant les capacités perceptives et mémorielles humaines au moyen de techniques interactives de visualisation. Nous proposons tout d'abord un système de navigation qui repose sur un fort couplage entre vues globales et vues locales pour aider à trouver les documents. Les vues globales montrent l'organisation de l'espace d'information et en permettent un survol rapide. Elles sont utilisées pour atteindre des zones d'intérêts. Les vues locales présentent le contenu de plusieurs documents en contexte afin d'aider l'utilisateur à les comparer et à choisir celui qu'il va consulter. Nous proposons ensuite un environnement visuel de gestion de bookmarks dans le but de pouvoir retrouver le plus simplement possible les documents préalablement consultés. Les utilisateurs organisent leurs bookmarks de la même manière qu'ils le feraient avec un ensemble de documents papiers sur un bureau. Ils définissent interactivement leur propre organisation conceptuelle de l'espace documentaire en se basant sur des propriétés visuelles et spatiales. Ces propriétés aideront les utilisateurs à retrouver plus efficacement les documents sauvegardés par une stimulation de leurs capacités mémorielles.
Llored, Yannick. "Etude herméneutique du texte goytisolien : au cours de la période 1980-1993." Lyon 2, 2002. http://theses.univ-lyon2.fr/documents/lyon2/2002/llored_y.
Full textThe purpose of this study is an analysis of the writing of Juan Goytisolo (Barcelona, 1931). It centres on the critical aspect of the writer's intellectual and literary project, and therefore on the way in which the sense of his texts is constructed in close relation to the Spanish literary tradition. It also presents a reconsideration of previous approaches to this writer's work, in order to examine, from a hermeneutic perspective, the work on language and the signification procedures developed by the author in accordance with precise objectives. The concepts of interpretation, reflexivity and reworking are thus at the centre of the demonstration, the purpose of which is to propose a new reading of this literary creation based on the renewal of aesthetic forms and on poetic expression which is open to the dialogue between cultures and traditions. The signification of the continuous movement of the repetitions which articulate Juan Goytisolo's writing technique, is compared to the thinking of the author and his view of the implications of the return to collective history and to cultural heritage, in order to question ideological discourses and the representations
Popovici, Eugen. "Recherche et filtrage d'information multimédia (texte, structure et séquence) dans des collections de documents XML hétérogènes." Phd thesis, Université de Bretagne Sud, 2008. http://tel.archives-ouvertes.fr/tel-00511981.
Full textElagouni, Khaoula. "Combinaison d'approches neuronales et de connaissances linguistiques pour la reconnaissance de texte dans les documents multimédias." Phd thesis, INSA de Rennes, 2013. http://tel.archives-ouvertes.fr/tel-00864923.
Full textMartin, Thomas. "Vers une reconnaissance multimodale du texte et de la parole pour l'analyse de documents vidéos pédagogiques." La Rochelle, 2009. http://www.theses.fr/2009LAROS264.
Full textThis work focuses on the implementation of methods for multimodal recognition of text and speech in audiovisual content. It focuses in particular on lecture records, in which text and speech are extensively used. As the production of multimedia data is massively increasing, the access to these data becomes problematic and requires an efficient content indexing. It is necessary to take into account the heterogeneous nature of this information which is the aim of the paradigm of multimodal analysis. It should however be noted that due to the recent emergence of multimodal analysis, there have been only few attempts to define this field. In addition, only few studies have focused on the interaction between text and speech in multimedia stream and the use of this interaction for their extraction. Our contribution focuses on two points. First, we wish to address the lack of definition by proposing a model of the multimodal analysis. Its goal is to propose a framework for a better description of applications using multimodal analysis, including clearly defining the concepts of modality and multimodality. The second point of our contribution is focused on the multimodal recognition of text and speech. We compare text and speech recognition processes then consider two cases of text-speech collaboration. The first one doesn't involve text recognition. It aims to improve speech recognition by using a thematized language model based on textual resources of the course. Despite the small size of this corpus, we show a significant improvement in recognition results. We are experiencing also a text and speech cross-recognition method based on the complementarity phonetic and written representations of language. We show that this approach improves the text recognition results and could be used to emphasize the specialized vocabulary of the course
Beneus-Frechard, Céline. "La médiation esthétique du texte : la modernité de la pensée du texte par l'introduction d'une problématique de la communication dans la réflexion sur l'expérience poétique, conçue comme expérience de la communication." Lyon 2, 2004. http://theses.univ-lyon2.fr/documents/lyon2/2004/beneus_c.
Full textTo think about an esthetic mediation consists in thinking the dialectic between the specificity of singularity, done at once, at the time of the creation of literature and at the time of the reading and the interpretation, and the specificity of the collectiv part, done at the time of the diffusion in the public space or the politic part (symbolic) or ideological (imaginary). These research work should account of the esthetic part of the media and therefore, confronting media and literature should explain the concept of information. We have to question the statute of the fiction in the thinking about the esthetic of communication, especially to show what the fiction brings about the communication thinking and the intelligibility of the facts of communication
Ghanmi, Nabil. "Segmentation d'images de documents manuscrits composites : application aux documents de chimie." Thesis, Université de Lorraine, 2016. http://www.theses.fr/2016LORR0109/document.
Full textThis thesis deals with chemistry document segmentation and structure analysis. This work aims to help chemists by providing the information on the experiments which have already been carried out. The documents are handwritten, heterogeneous and multi-writers. Although their physical structure is relatively simple, since it consists of a succession of three regions representing: the chemical formula of the experiment, a table of the used products and one or more text blocks describing the experimental procedure, several difficulties are encountered. In fact, the lines located at the region boundaries and the imperfections of the table layout make the separation task a real challenge. The proposed methodology takes into account these difficulties by performing segmentation at several levels and treating the region separation as a classification problem. First, the document image is segmented into linear structures using an appropriate horizontal smoothing. The horizontal threshold combined with a vertical overlapping tolerance favor the consolidation of fragmented elements of the formula without too merge the text. These linear structures are classified in text or graphic based on discriminant structural features. Then, the segmentation is continued on text lines to separate the rows of the table from the lines of the raw text locks. We proposed for this classification, a CRF model for determining the optimal labelling of the line sequence. The choice of this kind of model has been motivated by its ability to absorb the variability of lines and to exploit contextual information. For the segmentation of table into cells, we proposed a hybrid method that includes two levels of analysis: structural and syntactic. The first relies on the presence of graphic lines and the alignment of both text and spaces. The second tends to exploit the coherence of the cell content syntax. We proposed, in this context, a Recognition-based approach using contextual knowledge to detect the numeric fields present in the table. The thesis was carried out in the framework of CIFRE, in collaboration with the eNovalys campany.We have implemented and tested all the steps of the proposed system on a consequent dataset of chemistry documents
Moulin, Christophe. "Modélisation de documents combinant texte et image : application à la catégorisation et à la recherche d'information multimédia." Phd thesis, Université Jean Monnet - Saint-Etienne, 2011. http://tel.archives-ouvertes.fr/tel-00630438.
Full textMoysset, Bastien. "Détection, localisation et typage de texte dans des images de documents hétérogènes par Réseaux de Neurones Profonds." Thesis, Lyon, 2018. http://www.theses.fr/2018LYSEI044/document.
Full textBeing able to automatically read the texts written in documents, both printed and handwritten, makes it possible to access the information they convey. In order to realize full page text transcription, the detection and localization of the text lines is a crucial step. Traditional methods tend to use image processing based approaches, but they hardly generalize to very heterogeneous datasets. In this thesis, we propose to use a deep neural network based approach. We first propose a mono-dimensional segmentation of text paragraphs into lines that uses a technique inspired by the text recognition models. The connexionist temporal classification (CTC) method is used to implicitly align the sequences. Then, we propose a neural network that directly predicts the coordinates of the boxes bounding the text lines. Adding a confidence prediction to these hypothesis boxes enables to locate a varying number of objects. We propose to predict the objects locally in order to share the network parameters between the locations and to increase the number of different objects that each single box predictor sees during training. This compensates the rather small size of the available datasets. In order to recover the contextual information that carries knowledge on the document layout, we add multi-dimensional LSTM recurrent layers between the convolutional layers of our networks. We propose three full page text recognition strategies that tackle the need of high preciseness of the text line position predictions. We show on the heterogeneous Maurdor dataset how our methods perform on documents that can be printed or handwritten, in French, English or Arabic and we favourably compare to other state of the art methods. Visualizing the concepts learned by our neurons enables to underline the ability of the recurrent layers to convey the contextual information
Paganelli, Céline. "La recherche d'information dans des bases de documents techniques en texte integral. Etude de l'activite des utilisateurs." Grenoble 3, 1997. http://www.theses.fr/1997GRE39019.
Full textThis thesis is about information retrieval (ir) in full-text databases, and it particularly deals with information retrieval by expert users in technical documents. We suggest that the design of an ir system in technical document, should be approached by studying users' needs and expectations. We first present a state of art about ir and about the particularities of information retrieval in full-text. Then, experimental studies are led to approach the expert users' needs and expectations in case of information retrieval in technical documents, and a study of ir systems in full-text technical documents shows that these systems do not completely meet the users'needs. We so suggest that we should take the linguistic particularities of technical texts into account in order to improve ir in full technical documents. A psychological and linguistical study shows that a technical text is composed of two types of textual units. The first one describes and defines objects (object type) and the second one describes how to do an action (action type). These two types are validated by a cognitive study and present linguistical marks which can be automatically recognised. Both the study of users' needs and the study of linguistical characteristics of technical text, allow us to make propositions for the conception of a ir system in technical documents
Caro, Dambreville Stéphane. "Rôle des organisateurs paralinguistiques dans la consultation des documents électroniques." Grenoble 3, 1995. https://tel.archives-ouvertes.fr/tel-00451634.
Full textMétayer, Natacha. "Compréhension et stratégies d’exploration des documents pédagogiques illustrés." Thesis, Rennes 2, 2016. http://www.theses.fr/2016REN20001/document.
Full textIt is easier today to offer in one informative document many sources of information presented in different formats. Presenting various media can bring benefits in terms of learning performance, but dealing effectively with these different sources of information is complex. Therefore, offering documents guiding the learner when he is reading may be necessary to promote the construction of a qualitative mental model. Empirical studies that are conducted during this thesis endeavour to determine which formats are most effective while gradually increasing guidance within the document. Thus, four items are tested: the position of the picture relative to the text, the effect of the instructions, the text segmentation and the introduction of a guidance through a figure-ground contrast. Moreover, in order to bring new elements of reflection about how people explore an informative document and the impact of these strategies on performance, the eye movements of the learners were recorded. The results showed that changes in the format of information have an impact on the strategies of consultation of the document, including the increasing eye transitions between texts and illustrations. However, only the segmentation of the text in the form of semantic paragraphs brought benefits in terms of understanding
Quint, Vincent. "Une approche de l'édition structurée des documents." Phd thesis, Grenoble 1, 1987. http://tel.archives-ouvertes.fr/tel-00010612.
Full textl'organisation logique des composants du document. A partir de ce principe, on propose un méta-modèle
qui permet la description des structures logiques de toutes sortes de documents et de différents types
d'objets fréquents dans les documents : formules mathématiques, tableaux, schémas, etc... on associe aux
structures logiques des règles de présentation qui déterminent l'aspect graphique de leurs composants.
On montre l'intérêt de cette approche en présentant deux systèmes interactifs construits sur ce modèle :
l'éditeur de formules mathématiques Edimath et l'éditeur de documents Grif. La présentation de ces systèmes
s'appuie sur un état de l'art de la typographie informatique.
Soua, Mahmoud. "Extraction hybride et description structurelle de caractères pour une reconnaissance efficace de texte dans les documents hétérogènes scannés : Méthodes et Algorithmes parallèles." Thesis, Paris Est, 2016. http://www.theses.fr/2016PESC1069/document.
Full textThe Optical Character Recognition (OCR) is a process that converts text images into editable text documents. Today, these systems are widely used in the dematerialization applications such as mail sorting, bill management, etc. In this context, the aim of this thesis is to propose an OCR system that provides a better compromise between recognition rate and processing speed which allows to give a reliable and a real time documents dematerialization. To ensure its recognition, the text is firstly extracted from the background. Then, it is segmented into disjoint characters that are described based on their structural characteristics. Finally, the characters are recognized when comparing their descriptors with a predefined ones.The text extraction, based on binarization methods remains difficult in heterogeneous and scanned documents with a complex and noisy background where the text may be confused with a textured background or because of the noise. On the other hand, the description of characters, and the extraction of segments, are often complex using calculation of geometricaltransformations, polygon, including a large number of characteristics or gives low discrimination if the characteristics of the selected type are sensitive to variation of scale, style, etc. For this, we adapt our algorithms to the type of heterogeneous and scanned documents. We also provide a high discriminatiobn between characters that descriptionis based on the study of the structure of the characters according to their horizontal and vertical projections. To ensure real-time processing, we parallelise algorithms developed on the graphics processor (GPU). Our main contributions in our proposed OCR system are as follows:A new binarisation method for heterogeneous and scanned documents including text regions with complex or homogeneous background. In this method, an image analysis process is used followed by a classification of the document areas into images (text with a complex background) and text (text with a homogeneous background). For text regions is performed text extraction using a hybrid method based on classification algorithm Kmeans (CHK) that we have developed for this aim. This method combines local and global approaches. It improves the quality of separation text/background, while minimizing the amount of distortion for text extraction from the scanned document and noisy because of the process of digitization. The image areas are improved with Gamma Correction (CG) before applying HBK. According to our experiment, our text extraction method gives 98% of character recognition rate on heterogeneous scanned documents.A Unified Character Descriptor based on the study of the character structure. It employs a sufficient number of characteristics resulting from the unification of the descriptors of the horizontal and vertical projection of the characters for efficient discrimination. The advantage of this descriptor is both on its high performance and its simple computation. It supports the recognition of alphanumeric and multiscale characters. The proposed descriptor provides a character recognition 100% for a given Face-type and Font-size.Parallelization of the proposed character recognition system. The GPU graphics processor has been used as a platform of parallelization. Flexible and powerful, this architecture provides an effective solution for accelerating intensive image processing algorithms. Our implementation, combines coarse/fine-grained parallelization strategies to speed up the steps of the OCR chain. In addition, the CPU-GPU communication overheads are avoided and a good memory management is assured. The effectiveness of our implementation is validated through extensive experiments
Pekdağ, Bülent. "Influence des relations entre le texte et l'image d'un film de chimie sur l'activité cognitive d'un apprenant." Lyon 2, 2005. http://theses.univ-lyon2.fr/documents/lyon2/2005/pekdag_b.
Full textThis work deals with the construction of chemical movies and their use by students. The movies are related to acids and bases at the upper secondary school level. The theoretical framework of this research involves the action theory for the analysis of students' cognitive activity. As movies are concerned, the framework is adapted from a semiotic point of view. Chemical knowledge is categorised with the two world model involving perceptible and reconstructed levels. Research data were collected while students were using movies imbedded in a hypermedia. Movies had been created for the purpose of this work. The influence of the relation between text and picture over the activity of students has therefore been determined. Several factors have been found responsible for the choice of a movie by students: surface features, reformulations of a question, productions of new questions to understand information provided by a movie, non-remembering of specific information seen in movies. The latter seems in relation to the category of knowledge that is involved in the movie but not visible on the screen. We have also found the influence of cognitive load and double coding. Memorisation of information imbedded in the text of the film, but not displayed on the screen, is low. Information is best remembered when it comes from pictures than from text. Iconic representation helps students to reuse the knowledge. The effect of the text - picture relation over a learner's cognitive activity has also been analysed from a didactic point of view. Two kinds of movies have been compared, the P's, the text of which mainly describes the perceptible aspect of the picture and the R's, where the text interprets and reconstructs the picture. The R's are generally more and better used than the corresponding P's. The pictures of the latter are more used and the R's lead to a larger use of the specific vocabulary of chemistry. The copy - paste process is the most occurring way of using the P's and the R's text and picture
Max, Aurélien. "De la création de documents normalisés à la normalisation de documents en domaine contraint." Grenoble 1, 2003. http://www.theses.fr/2003GRE10227.
Full textWell-formedness conditions on documents in constrained domains are often hard to apply. An active research trend approaches the authoring of normalized documents through semantic specification, thereby facilitating such applications as multilingual production. However, the current systems are not able to analyse an existing document in order to normalize it. We therefore propose an approach that reuses the resources of such systems to recreate the semantic content of a document, from which a normalized textual version can be generated. This approach is based on two main paradigms : fuzzy inverted generation, which heuristically finds candidate semantic representations, and interactive negotiation, which allows an expert of the domain to progressively validate the semantic representation that corresponds to the original document
Goffin, Pascal. "An Exploration of Word-Scale Visualizations for Text Documents." Thesis, Université Paris-Saclay (ComUE), 2016. http://www.theses.fr/2016SACLS256/document.
Full textThis dissertation explores how embedding small data-driven contextual visualizations can complement text documents. More specifically, I identify and define important aspects and relevant research directions for the integration of small data-driven contextual visualizations into text. This integration should eventually become as fluid as writing and as usable as reading a text. I define word-scale visualisations as small data-driven contextual visualizations embedded in text documents. These visualizations can use various visual encodings including geographical maps, heat maps, pie charts, and more complex visualizations. They can appear at a range of word scales, including sizes larger than a letter, but smaller than a sentence or paragraph. Word-scale visualisations can help support and be used in many forms of written discourse such as text books, notes, blog posts, reports, stories, or poems. As graphical supplements to text, word-scale visualisations can be used to emphasize certain elements of a document (e.g. a word or a sentence), or to provide additional information. For example, a small stock chart can be embedded next to the name of a company to provide additional information about the past trends of its stocks. In another example, game statistics can be embedded next to the names of soccer teams or players in daily reports from the UEFA European Championship. These word-scale visualisations can then for example allow readers to make comparison between number of passes of teams and players. The main benefit of word-scale visualisations is that the reader can remain focused on the text as the visualization are within the text rather than alongside it.In the thesis, I make the following main contributions: I explore why word-scale visualisations can be useful and how to support their creation. I investigate placement options to embed word-scale visualisations and quantify their effects on the layout and flow of the text. As word-scale visualisations also have implications on the reader's reading behavior I propose a first study that investigates different word-scale visualisation positions on the reading behavior. I also explore how word-scale visualisations can be combined with interaction to support a more active reading by proposing interaction methods to collect, arrange and compare word-scale visualisations. Finally, I propose design considerations for the authoring of word-scale visualisations and conclude with application examples.In summary, this dissertation contributes to the understanding of small data-driven contextual visualizations embedded into text and their value for Information Visualization
Bossard, Aurélien. "Contribution au résumé automatique multi-documents." Phd thesis, Université Paris-Nord - Paris XIII, 2010. http://tel.archives-ouvertes.fr/tel-00573567.
Full textKeskes, Iskandar. "Discourse analysis of arabic documents and application to automatic summarization." Thesis, Toulouse 3, 2015. http://www.theses.fr/2015TOU30023/document.
Full textWithin a discourse, texts and conversations are not just a juxtaposition of words and sentences. They are rather organized in a structure in which discourse units are related to each other so as to ensure both discourse coherence and cohesion. Discourse structure has shown to be useful in many NLP applications including machine translation, natural language generation and language technology in general. The usefulness of discourse in NLP applications mainly depends on the availability of powerful discourse parsers. To build such parsers and improve their performances, several resources have been manually annotated with discourse information within different theoretical frameworks. Most available resources are in English. Recently, several efforts have been undertaken to develop manually annotated discourse information for other languages such as Chinese, German, Turkish, Spanish and Hindi. Surprisingly, discourse processing in Modern Standard Arabic (MSA) has received less attention despite the fact that MSA is a language with more than 422 million speakers in 22 countries. Computational processing of Arabic language has received a great attention in the literature for over twenty years. Several resources and tools have been built to deal with Arabic non concatenative morphology and Arabic syntax going from shallow to deep parsing. However, the field is still very vacant at the layer of discourse. As far as we know, the sole effort towards Arabic discourse processing was done in the Leeds Arabic Discourse Treebank that extends the Penn Discourse TreeBank model to MSA. In this thesis, we propose to go beyond the annotation of explicit relations that link adjacent units, by completely specifying the semantic scope of each discourse relation, making transparent an interpretation of the text that takes into account the semantic effects of discourse relations. In particular, we propose the first effort towards a semantically driven approach of Arabic texts following the Segmented Discourse Representation Theory (SDRT). Our main contributions are: A study of the feasibility of building a recursive and complete discourse structures of Arabic texts. In particular, we propose: An annotation scheme for the full discourse coverage of Arabic texts, in which each constituent is linked to other constituents. A document is then represented by an oriented acyclic graph, which captures explicit and implicit relations as well as complex discourse phenomena, such as long-distance attachments, long-distance discourse pop-ups and crossed dependencies. A novel discourse relation hierarchy. We study the rhetorical relations from a semantic point of view by focusing on their effect on meaning and not on how they are lexically triggered by discourse connectives that are often ambiguous, especially in Arabic. A thorough quantitative analysis (in terms of discourse connectives, relation frequencies, proportion of implicit relations, etc.) and qualitative analysis (inter-annotator agreements and error analysis) of the annotation campaign. An automatic discourse parser where we investigate both automatic segmentation of Arabic texts into elementary discourse units and automatic identification of explicit and implicit Arabic discourse relations. An application of our discourse parser to Arabic text summarization. We compare tree-based vs. graph-based discourse representations for producing indicative summaries and show that the full discourse coverage of a document is definitively a plus
Galani, Maria Elefthéria. "Entre l'analyse psychologique et l'analyse littéraire, la recherche du "noyau dur" du texte théâtral : un exemple : la dramaturgie de Jean Giraudoux." Lyon 2, 2001. http://theses.univ-lyon2.fr/documents/lyon2/2001/galani_me.
Full textThe present thesis intends to propose a new analysis of a dramatic text and to predict a possible representation. Our purpose consists in searching for the "hard core" of the text, between the literary analysis and the psychological analysis. The means that we try to discover the principal pivot of the play-text. Naturally, we do have the intention of limiting the comprehensive possibilities of the play. Futhermore, the stimuli of the director are not rejected ; in contrast, we make an effort to help him to start from a more solid base, to have some references which could be used as a reflexion track. We chose to use as a model the dramatic works of Jean Giraudoux. The interpretation proposed combines two different approaching methods. Firstly, we realise an analysis based on the stage directions in the form of tables : the plays are defined through the stage presence of the heros, the quantity of their text, the dramatic relations and the conflicting relations among the "dramatis personae". Secondly, we apply a stylistic analysis by constructing a corpus of the most fruit ful metaphors in the plays of Jean Giraudoux, organised into motives characters and fields. Finally, we provide every character of the plays with the metaphor which is identified with it and then, we attempt to reveal the key-character of every play following its stage directions as well as its most significant metaphor
Cheriat, Ahmed. "Une Méthode de correction de la structure de documents XML dans le cadre d'une validation incrémentale." Tours, 2006. http://www.theses.fr/2006TOUR4022.
Full textXML becomes the main tool used to exchange data on the web. In this context, XML document should respect schema constraints that describe the structural form of XML documents. The validation of an XML document wrt a schema constraints consists in testing whether the document verifies the set of structural specifications described by this schema. Supposing that updates are applied to the document, an incremental validator is the one that verifies whether the updated document complies with the schema, by validating only the parts of the document involved in the updates (to reduce the cost of a validation from scratch of the whole XML document). In this thesis we associate the validation process with correctoin proposals. During the execution of our validation method, if a constraint violation is found, a correction routine is called in order to propose local solutions capable of allowing the validation process to continue. Firstly, we are interested in a special case of this problem, which consists in correcting a word wrt another word (the initial valid word) and wrt a regular language. Indeed, the correction of an XML document having only a root and its sons corresponds to the correction of a word (composed by the children of the root) wrt a regular language (the constraint associated tothe root). In a second time, we extended this idea to the correction of XML tree (the structure of documents) wrt tree languages (the schema associated to an XML document). This correction is done by applying the minimum of modifications (by insertion, deletion or by replacement of some elements) on an invalid XML document in order to obtain a valid XML document. The algorithms presented in this thesis were implemented (in Java) and the experimental result are shown
Tarafdar, Arundhati. "Wordspotting from multilingual and stylistic documents." Thesis, Tours, 2017. http://www.theses.fr/2017TOUR4022/document.
Full textWord spotting in graphical documents is a very challenging task. To address such scenarios this thesis deals with developing a word spotting system dedicated to geographical documents with Bangla and English (Roman) scripts. In the proposed system, at first, text-graphics layers are separated using filtering, clustering and self-reinforcement through classifier. Additionally, instead of using binary decision we have used probabilistic measurement to represent the text components. Subsequently, in the text layer, character segmentation approach is applied using water-reservoir based method to extract individual character from the document. Then recognition of these isolated characters is done using rotation invariant feature, coupled with SVM classifier. Well recognized characters are then grouped based on their sizes. Initial spotting is started to find a query word among those groups of characters. In case if the system could spot a word partially due to any noise, SIFT is applied to identify missing portion of that partial spotting. Experimental results on Roman and Bangla scripts document images show that the method is feasible to spot a location in text labeled graphical documents. Experiments are done on an annotated dataset which was developed for this work. We have made this annotated dataset available publicly for other researchers
Trupin, Eric. "Segmentation de documents : Application a un systeme de lecture pour non-voyants." Rouen, 1993. http://www.theses.fr/1993ROUES009.
Full textPitou, Cynthia. "Extraction d'informations textuelles au sein de documents numérisés : cas des factures." Thesis, La Réunion, 2017. http://www.theses.fr/2017LARE0015.
Full textDocument processing is the transformation of a human understandable data in a computer system understandable format. Document analysis and understanding are the two phases of document processing. Considering a document containing lines, words and graphical objects such as logos, the analysis of such a document consists in extracting and isolating the words, lines and objects and then grouping them into blocks. The subsystem of document understanding builds relationships (to the right, left, above, below) between the blocks. A document processing system must be able to: locate textual information, identify if that information is relevant comparatively to other information contained in the document, extract that information in a computer system understandable format. For the realization of such a system, major difficulties arise from the variability of the documents characteristics, such as: the type (invoice, form, quotation, report, etc.), the layout (font, style, disposition), the language, the typography and the quality of scanning.This work is concerned with scanned documents, also known as document images. We are particularly interested in locating textual information in invoice images. Invoices are largely used and well regulated documents, but not unified. They contain mandatory information (invoice number, unique identifier of the issuing company, VAT amount, net amount, etc.) which, depending on the issuer, can take various locations in the document. The present work is in the framework of region-based textual information localization and extraction.First, we present a region-based method guided by quadtree decomposition. The principle of the method is to decompose the images of documents in four equals regions and each regions in four new regions and so on. Then, with a free optical character recognition (OCR) engine, we try to extract precise textual information in each region. A region containing a number of expected textual information is not decomposed further. Our method allows to determine accurately in document images, the regions containing text information that one wants to locate and retrieve quickly and efficiently.In another approach, we propose a textual information extraction model consisting in a set of prototype regions along with pathways for browsing through these prototype regions. The life cycle of the model comprises five steps:- Produce synthetic invoice data from real-world invoice images containing the textual information of interest, along with their spatial positions.- Partition the produced data.- Derive the prototype regions from the obtained partition clusters.- Derive pathways for browsing through the prototype regions, from the concept lattice of a suitably defined formal context.- Update incrementally the set of protype regions and the set of pathways, when one has to add additional data
Elagouni, Khaoula. "Combining neural-based approaches and linguistic knowledge for text recognition in multimedia documents." Thesis, Rennes, INSA, 2013. http://www.theses.fr/2013ISAR0013/document.
Full textThis thesis focuses on the recognition of textual clues in images and videos. In this context, OCR (optical character recognition) systems, able to recognize caption texts as well as natural scene texts captured anywhere in the environment have been designed. Novel approaches, robust to text variability (differentfonts, colors, sizes, etc.) and acquisition conditions (complex background, non uniform lighting, low resolution, etc.) have been proposed. In particular, two kinds of methods dedicated to text recognition are provided:- A segmentation-based approach that computes nonlinear separations between characters well adapted to the localmorphology of images;- Two segmentation-free approaches that integrate a multi-scale scanning scheme. The first one relies on a graph model, while the second one uses a particular connectionist recurrent model able to handle spatial constraints between characters.In addition to the originalities of each approach, two extra contributions of this work lie in the design of a character recognition method based on a neural classification model and the incorporation of some linguistic knowledge that enables to take into account the lexical context.The proposed OCR systems were tested and evaluated on two datasets: a caption texts video dataset and a natural scene texts dataset (namely the public database ICDAR 2003). Experiments have demonstrated the efficiency of our approaches and have permitted to compare their performances to those of state-of-the-art methods, highlighting their advantages and limits
Jiménez-Guarin, Claudia. "Opérations d'accès par le contenu à une base de documents textuels : application à un environnement de bureau." Grenoble INPG, 1989. http://tel.archives-ouvertes.fr/tel-00333333.
Full textMolinari, Gaëlle. "Représentation et acquisition de connaissances spécifiques à un domaine scientifique à partir de textes et d'illustrations : effet de la structure du domaine, des connaissances des lecteurs, des caractéristiques du texte et de ses illustrations." Lyon 2, 2002. http://theses.univ-lyon2.fr/documents/lyon2/2002/molinari_g.
Full textSorin, Laurent. "Contributions des architectures textuelles à l'accessibilité non-visuelle des documents numériques." Thesis, Toulouse 2, 2015. http://www.theses.fr/2015TOU20132/document.
Full textDigital accessibility plays a crucial role for the education, the social inclusion and the autonomy of impaired people. This work focused on a universal component of digital documents: text formatting. Colors, fonts and text disposition are far more than just an ornament; text formatting conveys important meaning for content comprehension, and allows reader to optimize their activity. For instance, a specific set of colors and font can be enough to indicate the presence of a title, which allows a global representation of the content themes. Thus, we aimed at making text formatting meaning accessible to visually impaired people, so they can obtain the same information as sighted readers, and also benefit from the same optimizations when accessing the document with synthetic voices
Felhi, Mehdi. "Document image segmentation : content categorization." Thesis, Université de Lorraine, 2014. http://www.theses.fr/2014LORR0109/document.
Full textIn this thesis I discuss the document image segmentation problem and I describe our new approaches for detecting and classifying document contents. First, I discuss our skew angle estimation approach. The aim of this approach is to develop an automatic approach able to estimate, with precision, the skew angle of text in document images. Our method is based on Maximum Gradient Difference (MGD) and R-signature. Then, I describe our second method based on Ridgelet transform.Our second contribution consists in a new hybrid page segmentation approach. I first describe our stroke-based descriptor that allows detecting text and line candidates using the skeleton of the binarized document image. Then, an active contour model is applied to segment the rest of the image into photo and background regions. Finally, text candidates are clustered using mean-shift analysis technique according to their corresponding sizes. The method is applied for segmenting scanned document images (newspapers and magazines) that contain text, lines and photo regions. Finally, I describe our stroke-based text extraction method. Our approach begins by extracting connected components and selecting text character candidates over the CIE LCH color space using the Histogram of Oriented Gradients (HOG) correlation coefficients in order to detect low contrasted regions. The text region candidates are clustered using two different approaches ; a depth first search approach over a graph, and a stable text line criterion. Finally, the resulted regions are refined by classifying the text line candidates into « text» and « non-text » regions using a Kernel Support Vector Machine K-SVM classifier
Ghorbel, Adam. "Generalized Haar-like filters for document analysis : application to word spotting and text extraction from comics." Thesis, La Rochelle, 2016. http://www.theses.fr/2016LAROS008/document.
Full textThe presented thesis follows two directions. The first one disposes a technique for text and graphic separation in comics. The second one points out a learning free segmentation free word spotting framework based on the query-by-string problem for manuscript documents. The two approaches are based on human perception characteristics. Indeed, they were inspired by several characteristics of human vision such as the Preattentive processing. These characteristics guide us to introduce two multi scale approaches for two different document analysis tasks which are text extraction from comics and word spotting in manuscript document. These two approaches are based on applying generalized Haar-like filters globally on each document image whatever its type. Describing and detailing the use of such features throughout this thesis, we offer the researches of document image analysis field a new line of research that has to be more explored in future. The two approaches are layout segmentation free and the generalized Haar-like filters are applied globally on the image. Moreover, no binarization step of the processed document is done in order to avoid losing data that may influence the accuracy of the two frameworks. Indeed, any learning step is performed. Thus, we avoid the process of extraction features a priori which will be performed automatically, taking into consideration the different characteristics of the documents
Ventresque, Anthony. "Espaces vectoriels sémantiques : enrichissement et interprétation de requêtes dans un système d'information distribué et hétérogène." Phd thesis, Université de Nantes, 2008. http://tel.archives-ouvertes.fr/tel-00457820.
Full textSegond, Luis Otavio. "Archivage électronique multimédia." Paris 6, 1986. http://www.theses.fr/1986PA066333.
Full text