Name
Abstract | ||
|---|---|---|
This paper introduces a Japanese question answering system called ASKMi (Answer Seeker/Knowledge Miner). We describe its system components with an emphasis on the Question Analyzer, which performs Sematic Role Analysis for answer type recognition, query expansion and document constraint generation. Semantic Role Analysis is a rule-based, language-independent technique for mapping textual fragments to arbitrary labels which we call semantic roles. To our knowledge, ASKMi is the first Japanese QA system that can perform query expansion and document constraint generation depending on what the answer types are. We are currently participating at the second Japanese QA track (NTCIR-4 QAC2), and this paper includes preliminary evaluations using the first standard Japanese QA test collection called QAC1. The Mean Reciprocal Rank (MRR) of ASKMi is currently around 0.5-0.7, which suggests that it is a competitive system. |
Year | Venue | Keywords |
|---|---|---|
2004 | RIAO | rule based,query expansion,question answering system,mean reciprocal rank |
Field | DocType | Citations |
Question answering,Constraint generation,Information retrieval,Query expansion,Computer science,Mean reciprocal rank,Semantic role labeling | Conference | 15 |
PageRank | References | Authors |
1.09 | 17 | 6 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Tetsuya Sakai | 1 | 1460 | 139.97 |
| Yoshimi Saito | 2 | 23 | 3.15 |
| Yumi Ichimura | 3 | 29 | 4.22 |
| Makoto Koyama | 4 | 83 | 8.49 |
| Tomoharu Kokubu | 5 | 31 | 4.68 |
| Toshihiko Manabe | 6 | 90 | 8.90 |