The automatic construction of faceted terminological feedback for interactive document retrieval
Peter Anick · Natural language processing · 2001
We present a linguistic method for the automatic construction of terminological feedback for use in interactive information retrieval. The method exploits the tendency for key domain concepts within result sets to participate in families of semantically related lexical compounds. We show how lexical dispersion may be used to identify terms that can serve in the role of “facets”, broad categories that organize domain concepts into separate hierarchies. The phrases in which these terms occur serve to provide a set of attributes, specializations, related concepts, etc., for the so-identified facet terms. We investigate the effect of a number of parameters on the output, with the aim of providing a relevant set of feedback categories while minimizing run-time overhead.