Answering Multiple Questions on a Topic From Heterogeneous Resources.
Boris Katz, Matthew W. Bilotti, Sue Felshin, Aaron marian Fernandes, Wesley Hildebrandt, Roni Katzir, Jimmy Lin, Daniel Loreto, Gregory A. Marton, Federico Mora, Özlem Uzuner · 2004
this paper we will call "topics" to dif- ferentiate them from the "focus" (which has been called the "target" of a factoid or list question) of each individual question within the topic. The primary tool we used for analyzing each question in the context of its topic was the START Natural Language Question Answering System (Katz et al., 2002; Katz, 1988; Katz, 1997). Three of its internal functions were exposed in a TREC-specific API, and enhanced to work with a wider array of questions: