摘要
中文问答系统通常由问题分析、信息检索、答案抽取组成。其中,问题分析中的问题的分类是否准确会直接关系到提取答案的准确度,所以在问答系统中起到关键性作用。本文主要介绍了中文问答系统的结构、问题分类体系以及方法,并提出结合基于规则的模式匹配与基于统计的机器学习的方法对问题进行分类,从而提高分类的准确度。
Chinese question answering system usually consists of question analysis,information retrieval and answer extraction.Among them,whether the classification of the question in the question analysis is accurate is directly related to the accuracy of extracting the answer,so it plays a key role in the question and answer system.This paper mainly introduces the structure,question classification system and method of Chinese question answering system,and proposes a classification based on rule-based pattern matching and statistical-based machine learning to improve the accuracy of classification.
作者
夏艳辉
聂百胜
胡金凤
XIA Yan-hui;NIE Bai-sheng;HU Jin-feng(China University of Mining & Technology (Beijing),Beijing 100083,China;Shijiazhuang Tiedao University,Shijiazhuang 050043,China;China Mobile Fuyang Branch,Fuyang 236000,China)
出处
《价值工程》
2019年第16期147-149,共3页
Value Engineering
基金
河北省社科基金(编号:HB17JY069)资助
关键词
开放域
中文问答系统
问题分类
open-domain
Chinese question answering system
question classification