网络问答资源挖掘综述  

Survey on Web Oriented Question-Answer Resource Mining

作　　者：王宝勋[1] 刘秉权[1] 孙承杰[1] 王晓龙[1] 

机构地区：[1]哈尔滨工业大学计算机科学与技术学院,哈尔滨150001

出　　处：《智能计算机与应用》2012年第6期54-58,共5页Intelligent Computer and Applications

基　　金：国家自然科学基金(61100094;61272383);高等学校博士学科点专项科研基金(20102302120053)

摘　　要：由大规模问答对组成的问答知识库对于自动问答技术的巨大支持作用是不言而喻的,从互联网获取知识并以问答对的形式加以保存是扩充和完善问答知识库的必要手段。阐述了面向互联网数据源的问答资源挖掘对于问答系统构建和自然语言处理研究的重要意义,概述了网络问答资源挖掘研究的主要任务以及面临的挑战,并通过对当前该领域的研究成果的综述分析了其中的重点研究问题。It is obvious that the question-answer knowledge base composed of large amounts of question-answer pairs（QA pairs） has played a significant role in supporting the automatic question-answering systems（QA systems）.Obtaining knowledge in the form of QA pairs from the web pages is of great necessity for enriching the knowledge base of QA systems.This paper has described the importance of the web oriented question-answer resource mining for QA system construction and natural language processing research,and summarizes the major tasks and challenge within this area.By surveying the research work,this paper has analyzed the essential research problems in this field.

关 键 词：问答资源挖掘 问答语义相关度计算 答案摘要 模式化答案抽取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]