ENSM-SE and UJM at INEX 2010: Scoring with Proximity and Tag Weights

Michel Beigbeder (1) , Mathias Géry (2) , Christine Largeron (2) , Howard Seck (3, 4)

1 G2I-ENSMSE - Centre de Génie Industriel et Informatique
2 LHC - Laboratoire Hubert Curien
3 CEREMADE - CEntre de REcherches en MAthématiques de la DEcision
4 Université Paris Dauphine-PSL

This paper presents our participation in the Relevant in Context task (ad-hoc track) during the 2010 INEX competition, and a posterior analysis. Two models presented in previous editions of INEX by the authors were merged for our 2010 participation. The first one is based on the proximity of the query terms in the documents and the second one is based on learnt tag weights. The results demonstrate the improvement of focused information retrieval, thanks to the integration of the tag weights in the approach based on proximity.

Keywords: XML Information Retrieval; Proximity Model; HTML Tags

Domains: Information Retrieval [cs.IR]

Michel Beigbeder, Mathias Géry, Christine Largeron, Howard Seck. ENSM-SE and UJM at INEX 2010: Scoring with Proximity and Tag Weights. 9th International Workshop of the Inititative for the Evaluation of XML Retrieval, INEX 2010, Dec 2010, Vught, Netherlands, France. pp.44-53. ⟨ujm-00634862⟩