Paper: Know-Why Extraction from Textual Data for Supporting What Questions

ACL ID W08-1603
Title Know-Why Extraction from Textual Data for Supporting What Questions
Venue Coling 2008: Proceedings of the workshop on Cross-Framework and Cross-Domain Parser Evaluation
Session
Year 2008
Authors

This research aims to automatically ex- tract Know-Why from documents on the website to contribute knowledge sources to support the question-answering sys- tem, especially What-Question, for dis- ease treatment. This paper is concerned about extracting Know-Why based on multiple EDUs (Elementary Discourse Units). There are two problems in ex- tracting Know-Why: an identification problem and an effect boundary determi- nation problem. We propose using Naïve Bayes with three verb features, a causa- tive-verb-phrase concept set, a supporting causative verb set, and the effect-verb- phrase concept set. The Know-Why ex- traction results show the success rate of 85.5% precision and 79.8% recall.