RESEARCH ON CHINESE RESTRICTED DOMAIN QUESTION ANSWERING SYSTEM

摘要：

The knowledge representing method, question semantic analysis and answer extraction algorithm are very important parts of QA (question answering system). This paper puts forward a framework which can well serve not only the simple question, but also the complex conditional question; can construct and apply ontology to drive Chinese restricted domain QA; and can conveniently extract answer by QSR (question semantic representation). Beside, the framework can automatically communicate with asker against askers fuzzy question. Firstly construct skillfully domain ontology; then analyze semantically askers question to get QSR; next automatically judge whether it is necessary to communicate with asker according to the ontology KB (knowledge base) and askers question; finally extract answer from domain ontology KB by OWLReasoner according to QSR. Through taking a consulting QA about computer maintenance as an example, we analyze its structure and implementation process, elaborate on construction of domain ontology, question semantic analysis and answer extraction algorithm. The experimental results are encouraging.

关键词： Restricted domain question answering system Ontology Question semantic analysis Knowledge representation

作者: JIE LIU JI-BIN FU XIAO-YU JIANG KANG CHEN

作者单位: Computer Science and Technology College, Beijing Institute of Technology, Beijing 100081, China Info Computer Science and Technology College, Beijing Institute of Technology, Beijing 100081, China

会议类型: 国际会议

会议名称: 2008 International Conference on Machine Learning and Cybernetics(2008机器学习与控制论国际会议)

会议地点: 昆明

会议语种:英文

页码: 2585-2590

在线出版日期: 2008-07-12（万方平台首次上网日期，不代表论文的发表时间）