In this paper, we check the limitations of image annotation on a large real corpus. We built a corpus of documents containing text and pictures, using more than 25000 press articles. We use this data to compare the similarity between a text-based retrieval and an image-based retrieval. The retrieval systems we use are common Information Retrieval (IR) tools. Results show that the two IR systems...

In this paper, we present and evaluate a multi-agent approach for polyhedral objectcontented range image segmentation. The approach consists in using a population of autonomous agents to segment a range image in its planar regions. The agents adapt to the regions on which they move, then perform cooperative and competitive actions allowing a collective segmentation of the image. The experimenta...

LINA at DEFT 2012 This article presents the participation of the TALN group at LINA to the défi fouille de textes (DEFT) 2012. Developed specifically for the second task, our system combines the outputs of three different keyword extraction methods. Our system ranked 2nd out of 9 systems with a f-measure of 21,3%. MOTS-CLÉS : extraction de mots clés, deft 2012, combinaison de méthodes.

Résumé. La recherche de documents similaires est un processus qui consiste à trouver les documents présentant des similitudes, comme la copie ou la reformulation, sur des bases documentaires ou sur internet. Elle est utilisée notamment pour protéger la propriété intellectuelle de productions issues de l’enseignement, de la recherche ou de l’industrie. Dans cet article, nous définissons une appr...

The birth of the XML standard and the growing use of images in electronic documents raised an open issue in information retrieval: image retrieval in semi-structured documents. This article presents a method to evaluate a semantic representation of images using the text and the document structure. More precisely, we propose a measure that evaluates the participation of each element of the docum...

Bayesian networks are currently one of the most interesting techniques of artificial intelligence. They combine the readability of a knowledge representation by an intuitive causal graph and the effectiveness of a data representation that takes into account the uncertainty in reasoning. They are used in various applications. This paper presents some ideas on the concept of Bayesian networks. We...

MOTS-CLÉS : objets, règles, contraintes, multimédia, traitement du signal.