Multimedia Information Modeling and Retrieval, Laboratoire d'Informatique de Grenoble (LIG-MRIM) at CHiC2013 - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2013

Multimedia Information Modeling and Retrieval, Laboratoire d'Informatique de Grenoble (LIG-MRIM) at CHiC2013

Résumé

Numerous cultural heritage materials are accessible through online digital library portals. However, this conversion resulted in the issues of inconsistency and incompleteness. The Cultural Heritage in CLEF 2013 (CHiC) takes the initiative to organize an evaluation campaign which involve several tasks such as 1) multilingual task, 2) polish task and 3) interactive task. We present the results of the MRIM/LIG team for the Ad-Hoc task and for the Semantic Enrichment task. For the Ad-Hoc task, we incorporate Term Links based on Wikipedia into the Language Model. Our approach has the following advantages: 1) it is easy and simple to generate the Term Similarity Matrix based on statistical information 2) a light weight integration in the Language Model. For the semantic query enrichment task, we deal with short queries found in this collection. These short queries cannot describe a specific information need. Hence, the goal of this task is to find best ten terms for a query to semantically enrich the topic and guess the user's information need or original query intent. We use the Wikipedia as a semantic resource in order to find these related terms.
Fichier non déposé

Dates et versions

hal-00953077 , version 1 (28-02-2014)

Identifiants

  • HAL Id : hal-00953077 , version 1

Citer

Kian Lam Tan, Mohannad Almasri, Philippe Mulhem, Jean-Pierre Chevallet, Catherine Berrut. Multimedia Information Modeling and Retrieval, Laboratoire d'Informatique de Grenoble (LIG-MRIM) at CHiC2013. CLEF'13, 2013, Unknown, Afghanistan. ⟨hal-00953077⟩
260 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More