An Empirical Study for the Automatic Acquisition of Topic Signatures
Montse Cuadros, Germán Rigau, Lluís Padró, S Omega · 2006
The main goal of this work is to compare different methods for building Topic Signatures, which are vectors of weighted words acquired from large corpora. We used two different software tools, ExRetriever [Cuadros et al., 2004] and Infomap [Dorow and Widdows, 2003], for acquiring Topic Signatures from corpus. Using these tools, we retrieve sense examples from large text collections. We also include in the comparison the Topic Signatures acquired previously by [Agirre and de la Calle, 2004] from the web. The three systems construct queries for each word sense using WordNet. ExRetriever and Infomap acquire the sense examples from the British National Corpus. The quality of the acquired Topic Signatures is indirectly evaluated on the Word Sense Disambiguation English Lexical Task of Senseval-2. 1