Ricoh Research at TREC 2006: Enterprise Track
Ganmei You, Yaojie Lu, Gang Li, Yueyan Yin · Text REtrieval Conference · 2006
1. Abstract This article presents our contributions to expert s earch and discussion search of Enterprise Track in TREC 2006. In discussion search, we take advantage of th e redundant patterns of emails, such as the subject , author, sent time, etc., which we incorporate in a field-based weighting method to mine discussion top ics with more robustness. Some non-content features, su ch as time-line and mail thread are found to be use ful as experiments showed they improve the precision of the search. In expert search, two variants of the BM25 and DFR_BM25 weighting models - namely V-BM25 and V-DFR_BM25 - are put forward. Query-based document length, not profile length, is used as document length in these weighting models to eliminate multi ple topic drift. In addition, we propose a variant of an existing phrase weighting model to decrease topic c onfusion (V-phrase) and a two-stage field-based sea rch method to refine the results. We demonstrate these approaches can effectively improve expert search.