Abstract
Question answering is an important field in artificial intelligence researches. But there are some limitations in traditional QA systems based on pattern matching. This paper analyzes and applies natural language process algorithms based on HMM, chart parsing, word dictionary and syntactic rules to extend dialogue management module, perform semantic analysis on users' sentences to implement semantic blocks recognition, theme recognition and information distillation. Those algorithms can improve system's process ability towards sentence analysis and overcome the disadvantages of traditional methods. The QA system is implemented on Java platform.
Original language | English |
---|---|
Pages (from-to) | 205-206 |
Number of pages | 2 |
Journal | Jisuanji Gongcheng/Computer Engineering |
Volume | 32 |
Issue number | 20 |
Publication status | Published - 20 Oct. 2006 |
Keywords
- HMM
- NLP
- Question answering
- Semantic analysis
- Syntax analysis