Bootstrap-Based Equivalent Pattern Learning for Collaborative Question Answering

Abstract

Semantically similar questions are submitted to collaborative question answering systems repeatedly even though these questions already contain best answers before. To solve the problem, we propose a precise approach of automatically finding an answer to such questions by identifying “equivalent” questions submitted and answered. Our method is based on a new pattern generation method T-IPG to automatically extract equivalent question patterns. Taking these patterns from training data as seed patterns, we further propose a bootstrap-based pattern learning method to extend more equivalent patterns on these seed patterns. The resulting patterns can be applied to match a new question to an equivalent one that has already been answered, and thus suggest potential answers automatically. We experimented with this approach over a large collection of more than 200,000 real questions drawn from Yahoo! Answers archive, automatically acquiring over 16,991 equivalent question patterns. These patterns allow our method to obtain over 57% recall and over 54% precision on suggesting an answer automatically to new questions, significantly improving over baseline methods.