摘要
知识图谱可以为智能问答和自动推荐等系统提供良好的数据支持。针对国内现有学科知识图谱构建数据来源单一等问题,提出一种多源异构数据融合的方法构建初中数学知识图谱。基于领域知识和学习者需求构建初中数学本体,确定概念、方法、公式、定理四种类型的实体;从教材等权威数据源和百度百科、互动百科等网络数据源中获取非结构化与半结构化数据,基于BERT(bidirectional encoder representations from transforms)模型抽取出教材等非结构化数据的关系和实体;利用基于层次过滤思想的知识融合模型进行多源异构数据的融合。实现了基于初中数学知识图谱的智能问答和自动推荐系统,为学习者提供及时且智能的学习支持服务,为破解初中数学在线教学个性化不足提供一条思路。
Knowledge graph can provide good data supporting for the intelligent question-answering system and automatic recommendation system. This paper proposes a method of multi-source heterogeneous data fusion to construct junior high school mathematics knowledge graph to solve the problem of a single data source of the existing subject knowledge graph in China. First, the junior high school mathematics ontology was constructed to determine four types of entities: concept, method, formula and theorem based on the domain knowledge and learners. needs. Then, unstructured and semi-structured data were obtained from authoritative data sources such as textbooks and network data sources such as Baidu Wikipedia,HDWiki, and the relationships and entities of unstructured data were extracted from the BERT(bidirectional encoder representations from transforms) model. Then the knowledge fusion model based on multi-pass sieve was used to fuse heterogeneous data from multiple sources. Finally, the intelligent question-answering and automatic recommendation system based on the knowledge graph of junior high school mathematics is realized. This study can provide learners with timely and intelligent learning support services, so as to provide an idea for solving the insufficiency of personalization in the online teaching of junior high school mathematics.
作者
周炫余
唐祯
唐丽蓉
李璇
卢笑
ZHOU Xuanyu;TANG Zhen;TANG Lirong;LI Xuan;LU Xiao(Key Laboratory of Big Data Research and Application for Basic Education,Hunan Normal Lniversity,Changsha 410006,Hunan,China;Institute of Education,Xiamen University,Xiamen 361005,Fujian.China;College of Engineering and Design,Hunan Normal University,Changsha 410006,Hunan,China;School of Computer Science,Wuhan University,Wuhan 430072,Hubei,China)
出处
《武汉大学学报（理学版）》
CAS
CSCD
北大核心
2021年第2期118-126,共9页
Journal of Wuhan University:Natural Science Edition
基金
国家自然科学基金(62007007、61703155)
湖南省自然科学基金(2018JJ3352、2018JJ3350)
国家级大学生创新创业训练计划支持项目(S201910542076)
湖南省大学生研究性学习和创新性实验计划项目(S201910542075)。
关键词
知识图谱
多源异构数据融合
层次过滤模型
初中数学
knowledge graph
multi-source heterogeneous data fusion
multi-pass sieve model
junior high school mathematics
