Overview of WebCLEF 2006

We report on the CLEF 2006 WebCLEF track devoted to crosslingual web retrieval. We provide details about the retrieval tasks, the used topic set, and the results of WebCLEF participants. WebCLEF 2006 used a stream of known-item topics consisting of: (i) manual topics (including a selection of WebCLEF 2005 topics, and a set of new topics) and (ii) automatically generated topics (generated using two techniques). Our main findings are the following. First, the results over all topics show that current CLIR systems are quite effective, retrieving on average the target page in the top few ranks. Second, when we break down the scores over the manually constructed and the generated topics, we see that the manually constructed topics result in higher performance. Third, the resulting scores on automatic topics give, at least, a solid indication of performance, and can hence be an attractive alternative in situations where manual topics are not readily available. Categories and Subject Descrip...

• Type: Conference
• Year: 2006
• Where: CLEF
• Authors: Krisztian Balog, Leif Azzopardi, Jaap Kamps, Maarten de Rijke
• Keywords: CLEF 2006, Generated Topics, Information Management, Manual Topics, Web Retrieval