Abstract:

Aiming at the inadequacy of ontology and transfer network in existing NEQAS applications，a method of chaptered ontology and transfer network is proposed.A confirming of question mode based on chaptered ontology and the making answers of standardization questions based on chaptered transfer network are presented.Semantic similarity model is discussed based on knowledge of chaptered ontology to solve the making answers of non-standardization questions，and the search method based on indexed transfer subnet is also built.Experimental results show that a method of chaptered ontology and transfer network can reduce the searching time by about 38% with the same efficiency as ontology and transfer network.

Key words: intelligent question answering system, ontology, transfer network, chapter

摘要： 针对现有本体和转移网络在网络教育自动答疑系统中应用的不足，提出了一种基于章节划分的本体与转移网络方法。定义了章节本体，确定了基于章节本体的问题模式，给出了基于章节划分转移网络解决规范化问题的答案生成。基于章节本体知识给出了语义相似度模型，用于解决非规范问题的答案生成；基于章节划分转移网络建立了章节索引转移子网的搜索方法。实验结果表明，基于章节划分的本体论和转移网络方法在不降低搜索效果的情况下，比一般的本体论和转移网络方法的查询时间减少了38%。

关键词: 智能答疑系统, 本体, 转移网络, 章节