Browsing Help for a Faster Retrieval

Browsing Help for Faster Document Retrieval Eric Crestan Sinequa, SinequaLabs 51-54, rue Ledru-Rollin 92400 Ivry-sur-Seine, France crestan@sinequa.com Laboratoire Informatique d’Avignon B.P. 1228 Agroparc 339 Chemin des Meinajaries 84911 Avignon Cedex 9, France Claude de LOUPY Sinequa, SinequaLabs 51-54, rue Ledru-Rollin 92400 Ivry-sur-Seine, France loupy@sinequa.com MoDyCo, Université de Paris 10 Laboratoire MoDyCo - UMR 7114 Université Paris 10, Bâtiment L 200, avenue de la République 92001 Nanterre Cedex, France Abstract In this paper, the search engine Intuition is described. It allows the user to navigate through the documents retrieved with a given query. Several “browse help” functions are provided by the engine and described here: conceptualisation, named entities, similar documents and entity visualization. They intend to “save the user’s time”. In order to evaluate the amount of time these features can save, an evaluation was made. It involves 6 users, 18 queries and the corpus is made of 16 years of the newspaper Le Monde. The results show that, with the different features, a user get faster to the needed information. fewer non-relevant documents are read (filtering) and more relevant documents are retrieved in less time. 1 Introduction During the last 10 years, TREC (Harman, 1993) allowed many researchers to evaluate their search engines and helped the field to progress. In 2000, Donna Harman studied the evolution of 2 search engines from 1993 (Harman, 2000). She showed that, after an improvement period, the performances have been almost the same for several years. This observation seems now classic: improving the heuristics or adding linguistic knowledge to a "good" engine does not dramatically improve its results. The problem is that even the best engines do not come up to the expectations of most users. So, if the performances do not really rise anymore, how can we rise users’ satisfaction? In fact, there are other ways to evaluate search engines than recall and precision. Time spent to find answers seems to be the most important one for the users and several papers present such an evaluation (Borlund and Ingwersen, 1998) (Järvelin and Kekäläinen, 2002). Considering the time factor, it is quite easy to improve the performances using procedures in order to help the user in his/her search. In this paper, we present the different 'browse help' features proposed to the users by Intuition, the search engine of Sinequa. First of all, we present the search engine itself (section 2). Then four types of help features are presented in section 3: conceptualisation, named entities filtering, similar documents and entity visualization. Section 4 describes the experiments done in order to evaluate the different browsing features and section 5 presents the results. These results show that using browsing help can decrease the time spent on searching. 2 Intuition search engine Intuition, the search engine of Sinequa, is based both on deep statistics and linguistic knowledge and treatments (Loupy et al., 2003). During indexing, the documents are analysed with a part of speech tagging, and a lemmatization procedure. But the most original linguistic feature of Intuition is the …

Content truncated to 500 words