当前位置>主页 > 期刊在线 > 信息技术 >

信息技术21年8期

基于约束型 LDA 的评价对象 - 情感词关联关系提取
万红新,彭欣悦
(江西科技师范大学 数学与计算机科学学院,江西 南昌 330038)

摘  要:随着社交媒体的发展,网络上产生了大量的评论类文本数据,这些数据蕴含了丰富的情感信息。对这些文本数据进行情感极性分类,需要提取评价对象 - 情感词的匹配关系。文章提出了基于约束 LDA 主题模型的评价对象 - 情感词关系提取模型 CE-LDA,将语义先验知识嵌入到 LDA 模型,在有效提取评价对象和情感词的同时,发现它们之间的关联性。通过实验分析,CE-LDA 模型对于评价对象和情感词及其关联关系的提取具有较好的效果。


关键词:语义知识;主题模型;约束嵌入;情感分类



DOI:10.19850/j.cnki.2096-4706.2021.08.008


基金项目:江西省高校人文社科项目(JC191 17);江西省教育厅科技项目(GJJ201127);江西 科技师范大学大学生创新创业训练计划项目(2021 11318002)


中图分类号:TP311                                     文献标识码:A                                     文章编号:2096-4706(2021)08-0027-03


Extraction of Association Relationship between Evaluation Object and Emotion Words Based on Constrained LDA

WAN Hongxin,PENG Xinyue

(School of Mathematics and Computer Science,Jiangxi Science & Technology Normal University,Nanchang 330038,China)

Abstract:With the development of social media,a large amount of comment text data has been generated on the internet,which contain rich emotion information. To classify the sentiment polarity of these text data,it is necessary to extract the matching relationship between the evaluation object and the emotion words. An the evaluation object and the emotion words relationship extracting model CELDA based on the constrained LDA topic model is proposed,which embeds semantic prior knowledge into the LDA model,and discovers the relevance between evaluation object and emotion words while effectively extracting them. Through experimental analysis,the CELDA model has a good effect on the extraction of evaluation object,emotion words and their associated relationships.

Keywords:semantic knowledge;topic model;constraint embedding;emotion classification


参考文献:

[1] BLEI D M,NG A Y,JORDAN M I. Latent Dirichlet Allocation [J].Journal of Machine Learning Research,2003,3(3): 993-1022.

[2] 谭旭,吴俊江,徐磊,等 . 面向深度学习的多维度中文 网络舆情分析 [J]. 小型微型计算机系统,2018,39(7):1471- 1477.

[3] TITOV I,MCDONALD R. Modeling online reviews with multi-grain topic models [C]//Proceedings of the 17th international conference on World Wide Web(WWW).New York:Association for Computing Machinery,2008:111-120.

[4] LIN C H,HE Y L. Joint sentiment/topic model for sentiment analysis [C]//Proceedings of the 18th ACM Conference on Information and Knowledge Management(ACM CIKM).New York:Association for Computing Machinery,2009:375-384.

[5] OZYURT B,AKCAYOL M A. A new topic modeling based approach for aspect extraction in aspect based sentiment analysis:SSLDA [J].Expert Systems with Applications,2020,168:114-231.

[6] THENMOZHI M,INDIRA R,DHARANI R. Using Lexicon and Random Forest Classifier for Twitter Sentiment Analysis [J]. International Journal of Computer Science and Engineering,2019,7 (6):591-594.

[7] ALBUQUERQUE P,VALLE D,LI D J. Bayesian LDA for mixed-membership clustering analysis:The Rlda package [J]. Knowledge-Based Systems,2019,163:988-995.

[8] 彭云,万常选,江腾蛟,等 . 基于语义约束 LDA 的商品 特征和情感词提取 [J]. 软件学报,2017,28(3):676-693.

[9] LU B,OTT M,CARDIE C,et al. Multi-aspect Sentiment Analysis With Topic Models [C]//Proceedings of the 11th IEEE International Conference on Data Mining Workshops(ICDMW).Vancouver:IEEE, 2011:81-88.

[10] PARK I,YOON B,KIM S,et al. Technological Opportunities Discovery for Safety Through Topic Modeling and Opinion Mining in the Fourth Industrial Revolution:The Case of Artificial Intelligence [J].IEEE Transactions on Engineering Management,2019, 99:1-16.

[11] 崔雪莲,那日萨,刘晓君 . 基于主题相似性的在线评论 情感分析 [J]. 系统管理学报,2018,27(5):821-827.


作者简介:万红新(1970—),女,汉族,江西南昌人,教授, 硕士,研究方向:数据挖掘、软件工程;彭欣悦(2001—),女, 汉族,江西宜春人,研究方向:软件工程、数据库技术。