查询结果:   陈志云,商月,钱冬明.基于知识图谱的智能答疑系统研究[J].计算机应用与软件,2018,35(2):178 - 182.
中文标题
基于知识图谱的智能答疑系统研究
发表栏目
人工智能与识别
摘要点击数
1171
英文标题
RESEARCH ON INTELLIGENT QUESTION ANSWERING SYSTEM BASED ON KNOWLEDGE GRAPH
作 者
陈志云 商月 钱冬明 Chen Zhiyun Shang Yue Qian Dongming
作者单位
华东师范大学 上海 200062     
英文单位
East China Normal University,Shanghai 200062,China     
关键词
中文知识图谱 智能答疑 朴素贝叶斯 问题分类
Keywords
Chinese knowledge graph Question answering Naive bayes Question classification
基金项目
国家高技术研究发展计划项目(2013AA01A211)
作者资料
陈志云,副教授,主研领域:多媒体技术及应用。商月,硕士生。钱冬明,副教授。 。
文章摘要
在大型开放式网络课程(MOOC)系统中,答疑是一个重要环节,智能答疑系统的研发,有助于提高教学的效能。一般的答疑系统缺乏对课程知识的智能表示,对以往问题答案的自动归纳等。利用知识图谱技术将学生提问问题以知识点树的形式显示,同时结合朴素贝叶斯算法文本分类,将问题关键字按照知识点树中的知识点进行归类,并与教学案例课件进行分类匹配,从而初步改造为智能答疑系统。系统以华东师范大学公共计算机课为应用实例,初步应用测试问题分类成功率均在80%左右,证明智能答疑系统能更好地表示和分类问题,更好地适应学生的答疑应用。
Abstract
In the massive open online courses (MOOC) system, the question answering is an important link. The development of the intelligent question answering system is helpful to improve the teaching efficiency. The general question answering system is lack of the intelligent representation of the course knowledge, and the automatic induction of the previous questions, and so on. This paper used Knowledge Graph technology to represent questions in the form of knowledge point tree. At the same time, this paper classified problems according to knowledge points in the knowledge tree, and match problems with teaching case courseware by the Naive Bayes algorithm. The system, which initially transformed into intelligent answering system, took the public computer course of East China Normal University as an application example. After testing four classes of students studied Data Processing and Management, we found that classification success rate can be about 80%, which proved that intelligent answering system could represent and classified problems better, thus being suitable to Q & A for students.
下载PDF全文