摘要
[目的/意义]在线教育环境下,用户基于MOOC平台生成了大量UGC笔记,通过UGC知识融合方式实现MOOC课程笔记的自动生成,有助于缓解大规模数据下的信息过载问题。[方法 /过程]在UGC数据抓取的基础上,针对UGC的非理性特点,采用基于聚类分析的融合方法对UGC笔记进行单条笔记融合,并依据加权策略抽取最优单条笔记,根据课程框架体系填充若干最优单条笔记,最终生成完整、准确的课程笔记。[结果/结论]结果表明,文章所构建的基于UGC知识融合的课程笔记自动生成模型对于MOOC课程笔记生成具有较好的效果,准确率、召回率以及F1值均达到了90.00%。
[Purpose/significance]Under the environment of online education,users generate a large number of notes based on the MOOC platform.The automatic generation of MOOC course notes is realized by means of UGC(user generated notes) knowledge fusion,which is conducive to alleviating the problem of information overload under large-scale data.[Method/process]On the basis of UGC data capture,this paper analyzes the irrational characteristics of UGC,and fuses UGC notes to generate single notes based on clustering analysis,and then extracts the optimal single notes according to weight strategy,fills the course framework system with them.Finally,the complete and accurate course notes are generated.[Result/conclusion]The results show that,the automatic generation model of course notes based on UGC knowledge fusion has a good effect on the generation of MOOC course notes,with the precision,recall and F1-measure reaching 90.00%.
引文
[1]RUSH A M,CHOPRA S,WESTON J.A neural attention model for abstractive sentence summarization[C]//Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing.Boston:The MIT Press,2015:379-389.
[2]NALLAPATI R,ZHOU B,SANTOS C D,et al.Abstractive text summarization using sequence-to-sequence RNNs and beyond[C]//Proceedings of the 20th SIGNLL Conference on Computational Natural Language Learning.Boston:The MITPress,2016:280-290.
[3]贾晓婷,王名扬,曹宇.结合Doc2Vec与改进聚类算法的中文单文档自动摘要方法研究[J].数据分析与知识发现,2018(2):86-95.
[4]CHAKRABARTI D,PUNERA K.Event summarization using tweets[C]//Proceedings of the Fifth International AAAIConference on Weblogs and Social Media.Menlo Park:The AAAI Press,2011:66-73.
[5]唐晓波,朱娟.大数据环境下知识融合的关键问题研究综述[J].图书馆杂志,2017,36(7):10-16.
[6]林海伦,王元卓,贾岩涛,等.面向网络大数据的知识融合方法综述[J].计算机学报,2017,40(1):1-27.
[7]李培.信息检索与信息融合[M].北京:人民邮电出版社,2007,46-52.
[8]SANTOS E,WILKINSON J T,SANTOS E E.Fusing multiple Bayesian knowledge sources[J].International Journal of Approximate Reasoning,2011,52(7):935-947.
[9]GOU J,YANG J G,CHEN Q.Evolution and evaluation in knowledge fusion system[M].Artificial Intelligence and Knowledge Engineering Applications:A Bioinspired Approach.Berlin:Springer,2005:192-201.
[10]黄新平.政府网站信息资源多维语义知识融合研究[D].长春:吉林大学,2017.
[11]KRIEGEL E U,MEISSENU,PFENNIGSCHMIDT S,et al.A knowledge fusion toolkit for decision making[C]//Third International Conference on Emerging Security Technologies.Piscataway:IEEE,2012:90-93.
[12]HU S K,GAO Y D.Web text knowledge fusion[C]//Second Pacific-Asia Conference on Web Mining and Web-based Application.Piscataway:IEEE,2009:171-174.
[13]沈旺,李亚峰,侯昊辰.数字参考咨询知识融合框架研究[J].图书情报工作,2013(19):139-143.
[14]唐晓波,刘广超.基于两层知识融合的金融知识服务模型研究[J].图书馆学研究,2018(16):79-85.
[15]周利琴,范昊,潘建鹏.网络大数据中的知识融合框架研究[J].情报杂志,2018,37(1):145-150,197.
[16]林鑫,梁宇.用户社会化标注中非理性行为的表现及原因分析[J].数字图书馆论坛,2016(12):48-53.
[17]罗睿阳.推荐算法研究和对象搜索引擎的搭建[D].北京:北京邮电大学,2013.
[18]束珏,成卫青,邓聪.基于话题标签和转发关系的微博聚类和主题词提取[J].计算机应用,2016,36(2):460-464.
[19]丁晟春,王楠,吴靓婵媛.基于关键词共现和社区发现的微博热点主题识别研究[J].现代情报,2018,38(3):10-18.
[20]李琳,李辉.一种基于概念向量空间的文本相似度计算方法[J].数据分析与知识发现,2018(5):48-58.
[21]文本相似度在线计算[EB/OL].[2018-06-19].http://textsimilarity.51tools.info/.