Use of Language Model, Phrases and Wikipedia Forward Links for INEX 2009

摘要

We present in this paper the work of the Information Retrieval Modeling Group (MRIM) of the Computer Science Laboratory of Grenoble (LIG) at the INEX 2009 Ad Hoc Track. Our aim this year was to twofold: first study the impact of extracted noun phrases taken in addition to words as terms, and second using forward links present in Wikipedia to expand queries. For the retrieval, we use a language model with Dirichlet smoothing on documents and/or doxels, and using an Fetch and Browse approach we select rank the results. Our best runs according to doxel evaluation get the first rank on the Thorough task, and according to the document evaluation we get the first rank for the Focused, Relevance in Context and Best in Context tasks.