Using Longest Common Subsequence Matching for Chinese Information Retrieval

This paper is about adopting the longest common subsequence (LCS) matching for Chinese information retrieval. We re-ranked the retrieved documents by a mixture of the original similarity score and the LCS score obtained by matching the document titles and the query. This LCS-based similarity score is also used in pseudo-relevance feedback in various ways (e.g., selecting terms and filtering documents with low LCS values). We evaluated the use of LCS in title re-ranking and PRF based on the NTCIR-4 test collection for Chinese ad hoc information retrieval. For title queries, our best MAP achieved is 26.7% evaluated using rigid relevance judgement and 30.2% evaluated using relax relevance judgement.