Our study compares the effectiveness of an information retrieval system based on the proximity of the query term occurrences in the documents and an IRS based on a language model with Dirichlet smoothing and with the Okapi BM25 model. Our proximity based model computes at each position in the document a value much higher as some occurrences of all the query terms are close to this position. Mor...

This article presents our contributions to expert search and discussion search of Enterprise Track in TREC 2006. In discussion search, we take advantage of the redundant patterns of emails, such as the subject, author, sent time, etc., which we incorporate in a field-based weighting method to mine discussion topics with more robustness. Some non-content features, such as time-line and mail thre...

Feng Guan, Xiaoming Yu, Zeying Peng, Hongbo Xu, Yue Liu, Linhai Song,Xueqi Cheng 1. Institute of Computing Technology, Chinese Academy of Sciences, Beijing, 100190 2. Graduate School of Chinese Academy of Sciences, Beijing 100190 Abstract This paper is about the work done for ad-hoc task of TREC 2009 Web Track. We introduce three methods for this task, including two improved BM25 models and que...

In this article we show how power transformations can be used as a common framework for the derivation of local term weights. We found that under some parametric conditions, BM25 and inverse regression produce equivalent results. As a special case of inverse regression, we show that the largest increment in term weight occurs when a term is mentioned for the second time. A model based on invers...

This paper describes the method we used for the Genomics Track of TREC 2006. BM25 model is implemented to retrieve relevant documents. We also tried to re-ranking documents based on the initial retrieval before passage retrieval. Passages are retrieved based on the concepts defining in topics and concept coverage. Results of submitted runs are listed and discussed.

This is the first year for the participation of the City University Centre of Interactive System Research (CISR) in the Expert Search Task. In this paper, we describe an expert search experiment based on windowbased techniques, that is, we build profile for each expert by using information around the expert’s name and email address in the documents. We then use the traditional IR techniques to ...

By the time of submission deadline, we completed two tasks: known-item search and discussion search. For both tasks, we used the PADRE retrieval system [1], in which the Okapi BM25 relevance function was implemented. Each message in the collection was treated as an independent document, so both topic distillation scoring and same site suppression mechanism were turned off (i.e. -nocool and –SSS...

RÉSUMÉ. L'objet de ce papier est de montrer qu'il est possible de mettre le modèle vectoriel et la pondération tf*idf associée en correspondance avec le calcul d'une résonance dans un réseau associatif basé sur une méthode très simple de propagation d'activation. Nous décrivons un réseau associatif associant termes et documents puis un calcul de résonance entre une requête et un document dans c...