Learning Web Query Patterns for Imitating Wikipedia Articles

This paper presents a novel method for acquiring a set of query patterns to retrieve documents containing important information about an entity. Given an existing Wikipedia category that contains the target entity, we extract and select a small set of query patterns by presuming that formulating search queries with these patterns optimizes the overall precision and coverage of the returned Web information. We model this optimization problem as a weighted maximum satisfiability (weighted Max-SAT) problem. The experimental results demonstrate that the proposed method outperforms other methods based on statistical measures such as frequency and point-wise mutual information (PMI), which are widely used in relation extraction.

• Type: Journal
• Year: 2010
• Where: COLING
• Authors: Shohei Tanaka, Naoaki Okazaki, Mitsuru Ishizuka
• Keywords: COLING 2010, Computational Linguistics, Point-wise Mutual Information, Query Patterns, Weighted Maximum Satisfiability