Research of Question Analysis Based on HNC and K Nearest Neighbor
The question answering system is a kind of information retrieval system, which accepts natural language questions and then finds out the accurate and concise answers from a large number of heterogeneous data. The question answering system can provide the user really useful, accurate information, rather than return thousands of documents links as the most current keyword-based string matching information retrieval systems do. To improve the accuracy of the question answering system, the restricted-domain question answering system is becoming a research Hotspot. In essence, the point to realize a restricted-domain question answering system is the understanding of user's questions and the domain knowledge text, so that the system can extract the information from knowledge text according to the specific question and then converts it to user requirement.