Dans cet article, nous proposons une méthode d’appariement de contenus d’actualité multimédias, considérant les exigences à la fois sémantiques et temporelles du besoin d’information. La pertinence d’une vidéo pour un article de presse est mesurée par deux indices, l’un saisissant la similarité de leurs contenus, l’autre la cohérence de leurs dates d’édition. Nous présentons également une méthodologie d’évaluation s’affranchissant des standards comparant les résultats du système à des résultats de référence, en soumettant les paires de documents proposées automatiquement à un panel d’utilisateurs chargé de juger de leur pertinence.

The linguistic resources presented in this paper are designed for the recognition and semantic tagging of calendar expressions in French. While existing resources generally put the emphasis on describing calendar bases pointed out by calendar expressions (which are considered as named entities), our approach tries to explicit how references to calendar are linguistically built up, taking into account not only the calendar bases but as well the prepositions and units that operate on them, as they provide valuable information on how texts refer to the calendar. The modelling of these expressions led us to consider calendar expressions as a conjunction of operators interacting with temporal references. Though the resources aim to be generic and easily reusable, we illustrate the interest of our approach by using the resources output to feed a text navigation tool that is currently being improved, in order to offer users a way of temporally progressing or navigating in texts.

This paper presents our work on the detection of temporal information in web pages. The pages examined within the scope of this study were taken from the tourism sector and the temporal information in question is thus particular to this area. The differences that exist between extraction from plain textual data and extraction from the web are brought to light. These differences mainly concern the spatial arrangement of the text, the use of punctuation and the respect of traditional syntactic rules. The temporal expressions to be extracted are classified into two kinds: temporal information that concerns one particular event and repetitive temporal information. We adopt a symbolic approach relying on patterns and rules for the detection, extraction and annotation of temporal expressions; our method is based on the use of transducers. First evaluations have shown promising results. Since the visual structure of a web page is very important and often informs the user before he has even read the text, a semiotic study is also presented in this paper.