Brigitte Simonnot

Cooperation model for video-document retrieval

Information retrieval in video document archives presents specific issues. One of them is that a video document contains a large amount of information and can be seen under different aspects. We propose an information retrieval process model based on the cooperation between different specialists: specialists in the application domain, specialists of the media, and specialists in information retrieval. Each specialist has a proper point of view on documents, a partial knowledge which can be exploited in the query interpretation and during the search, and a particular role to play in the different stages of the retrieval process. A facetted data model helps to refine documents descriptions and search results. Each facet can be linked to one structure level of video documents. During the retrieval process, a flexible collaboration between several information retrieval experts is set up to deal with the different aspects of documents and query descriptions and to improve retrieval performance. A prototype, using the MHEG standard, is being implemented to retrieve TV news sequences and to present search results in a hypermedia form.

Model for interactive retrieval of videos and still images

No single mechanism can be expected to give optimal retrieval in all cases and many problems have been reported concerning the difficulty of making a unique choice at the design stage of information retrieval systems. The paper presents a flexible model for the information retrieval process, which provides relevance feedback facilities and embodies a mechanism to exploit a typology of information needs. Various alternatives for information retrieval primitives are integrated in a single framework. We study the adequacy of this model to the retrieval of multimedia documents, namely pictures and videos. Multimedia documents and queries are commonly represented by a simple list of attribute-descriptor pairs. Introducing a facet level in descriptions allows to give a more expressive role to each keyword. This rather simple model is nevertheless well exploited by the proposed IR model which takes benefits from the faceted structure all along the search process.

Place des moteurs de recherche dans l’éditorialisation du web

En vingt ans, les moteurs web ont conquis une place considerable dans le paysage informationnel et le referencement aupres d’eux est devenu une activite a part entiere. Objets de contestations, ils se sont longtemps abrites derriere une neutralite, voire une objectivite censee etre garantie par un traitement informatise. Des controverses les ont pousses a clarifier leurs politiques editoriales. Analyser les consignes que Google et Bing donnent aux webmestres permet de relever des positionnements differents de ces deux firmes. En outre, on s’est moins interesse aux cas de dereferencement. La decision de la CJUE en 2014, dite du droit a l’oubli, en fournit l’occasion. Les reponses qu’y ont apportees les moteurs eclairent certains aspects de leurs politiques mais invitent aussi a reconceptualiser la notion d’editorialisation.