Cheshire at GeoCLEF 2007: Retesting Text Retrieval Baselines

In this paper we will briefly describe the approaches taken by Berkeley for the main GeoCLEF 2007 tasks (Mono and Bilingual retrieval). This year we used only a single system in the research, and were not able to do much in the way of interesting geographic work due to a number of factors, not the least of which was time competition from other tasks. The approach this year used probabilistic text retrieval based on logistic regression and incorporating blind relevance feedback for all of the runs. All translation for bilingual tasks was performed using the LEC Power Translator PCbased MT system. Since data on overall performance relative to other participants were not available at the time of writing, our discussion in this paper is limited to comparison between our submitted runs. Categories and Subject Descriptors H.3 [Information Storage and Retrieval]: H.3.1 Content Analysis and Indexing; H.3.3 Information Search and Retrieval General Terms Algorithms, Performance, Measurement Ke...

• Type: Conference
• Year: 2007
• Where: CLEF
• Authors: Ray R. Larson
• Keywords: Bilingual Retrieval, CLEF 2007, GeoCLEF 2007 Tasks, Information Retrieval, Logistic Regression