基于社交网络大数据的民众情感监测研究

doi:10.11959/j.issn.2096-0271.2022054

摘要/Abstract

摘要：

近年来，新浪微博、Twitter等社交网络平台逐渐成为反映社会舆情的主要载体之一，为网民发表观点和表达情绪提供了一个便利的平台。一旦突发事件发生，民众借助社交网络平台发布微博、推文等来表达与此相关的态度，这些信息通过社交网络进一步被传播扩散，从而产生一定的社会影响。基于社交网络大数据的舆情监控已经成为新的研究热点，利用各国的社交网络大数据进行民众情感监测，有助于直接掌握国际关系中的民众情感倾向，对我国外交、政治、对外贸易等方面都有很重要的作用。基于此，提出了一种面向中日语料的民众情感监测系统，该系统能够同时分析微博和Twitter等社交平台的中日文数据中包含的情感倾向，并以可视化的形式展现给用户。在情感分析算法上，在BERT模型基础上结合自扩展的中日文情感词典，提出了一个新的情感分类模型—EmoBERT。实验结果表明，相比于原始BERT模型，提出的EmoBERT模型在中文情感分类任务和日文情感分类任务上都取得了很好的效果。其中EmoBERT-C将中文BERT准确率从89.68%提升至92.15%，EmoBERT-J将日文BERT模型准确率从74.73%提升至78.26%。

关键词:

情感分析, 舆情监测, 情感词典, 中日关系, 微博, Twitter

Abstract:

In recent years, social networking platforms such as Sina Weibo and Twitter have gradually become one of the main carriers for reflecting social public opinion, providing a convenient platform for netizens to express their opinions and express their emotions. Once an emergency occurs, people use social network platforms to publish microblogs, tweets, etc. to express their attitudes related to this. These information are further spread and spread through social networks, thus producing a certain social impact. Public opinion monitoring based on social network big data has become a new research hotspot. The use of social network big data in various countries to monitor people's emotions is helpful to directly grasp people's emotional tendencies in international relations, and has a great impact on my country's diplomacy, politics, foreign trade and other aspects. have important roles. Based on this, this paper proposes a public sentiment monitoring system for Chinese and Japanese data, which can simultaneously analyze the emotional tendencies contained in Chinese and Japanese data on social platforms such as Weibo and Twitter, and display it to users in a visual form. In the sentiment analysis algorithm, based on the BERT model and combined with the self-expanding Chinese and Japanese sentiment dictionary, we propose a new sentiment classification model—EmoBERT. The experimental results show that, compared with the original BERT model, the EmoBERT proposed in this paper has achieved good results on both Chinese sentiment classification tasks and Japanese sentiment classification tasks. Among them, EmoBERT-C increased the accuracy of Chinese BERT from 89.68% to 92.15%, and EmoBERT-J increased the accuracy of Japanese BERT model from 74.73% to 78.26%.

Key words: Sentiment analysis, Public opinion monitor, Sentiment lexicon, Chinese-Japanese relations, Weibo, Twitter

李爱黎, 张子帅, 林荫, 王秋菊, 杨建安, 孟炜程, 张岩峰.

基于社交网络大数据的民众情感监测研究 [J]. 大数据, doi: 10.11959/j.issn.2096-0271.2022054.

LI Aili, ZHANG Zishuai, LIN Yin, WANG Qiuju, YANG Jianan, MENG Weicheng, ZHANG Yanfeng.

Research on Emotion Monitoring of Chinese-Japanese Public Based on Social Network Big Data [J]. Big Data Research, doi: 10.11959/j.issn.2096-0271.2022054.

参考文献

[1] 敦欣卉，张云秋，杨铠西. 基于微博的细粒度情感分析[J]. 数据分析与知识发现，2017， 001(007)：61-72.

Dun X, Zhang Y, Yang K. Fine-grained Sentiment Analysis Based on Microblog[J]. Data Analysis and Knowledge Discovery, 2017, 001(007): 61-72.

[2] Zhao J, Dong L, Wu J, et al. MoodLens: An emoticon-based sentiment analysis system for chinese tweets. ACM, 2012.

[3] Wang H, Can D, Kazemzadeh A, et al. A System for Real-time Twitter Sentiment Analysis of 2012 U.S. Presidential Election Cycle[C]// Acl System Demonstrations. 2012.

[4] Williams J, Katz G. Extracting and modeling durations for habits and events from Twitter[C]// Meeting of the Association for Computational Linguistics: Short Papers. 2013.

[5] 李忠俊. 基于话题检测与聚类的内部舆情监测系统[J]. 计算机科学，2012，39(012)：237-240.

LI Z J. Internal Public Opinion Monitoring System Based on Topic Detection and Clustering [J]. Computer Science, 2012, 39(012): 237-240.

[6] Yi J, Nasukawa T, Bunescu R, et al. Sentiment analyzer: extracting sentiments about a given topic using natural language processing techniques[C]// Third IEEE International Conference on Data Mining. IEEE, 2003.

[7] Riloff E M, Shepherd J. A Corpus-Based Approach for Building Semantic Lexicons[J]. 1997.

[8] 熊德兰，程菊明，田胜利. 基于HowNet的句子褒贬倾向性研究[J]. 计算机工程与应用， 2008，(22)：143-145.

Xiong D, Cheng J, Tian Shengli. A Study of Sentence Praisal and Derogation Tendency Based on HowNet [J]. Computer Engineering and Applications, 2008, (22): 143-145.

[9] 潘明慧，牛耘. 基于多线索混合词典的微博情绪识别[J]. 计算机技术与发展，2014，(9)： 28-32.

Pan M H, Niu Y. Microblog Emotion Recognition Based on Multi-cue Hybrid Dictionary [J]. Computer Technology and Development, 2014, (9): 28-32.

[10] Pang B. Thumbs up Sentiment Classification Using Machine Learning Techniques[J]. Proc. EMNLP, Philadelphia. PA, USA, 2002.

[11] 杨艳霞. 基于分类的微博情感分析算法研究及实现[J]. 计算机与数字工程，2017，45(2)： 197-197.

Yang Y X. Microblog Sentiment Analysis Algorithm Research and Implementation Based on Classification [J]. Computer & Digital Engineering, 2017, 45(2): 197-200.

[12] COLLOBERT R, WESTON J, BOTTOU L, et al. Natural language processing(almost)from scratch[J]. The Journal of Machine Learning Research, 2011, 12:2493-2537.

[13] Al-Rifaie M M, Bishop J M. Swarmic Sketches and Attention Mechanism[J]. Springer, Berlin, Heidelberg, 2013, 85–96,.

[14] Mousa E D, Vryniotis V. Sentiment analysis and opinion mining: on optimal parameters and performances[M]. John Wiley & Sons, Inc. 2015.

[15] 宋婷，陈战伟，杨海峰. 基于分层注意力网络的方面情感分析[J]. 大数据，2020，6(5)：10.

Ting SONG, Zhanwei CHEN, Haifeng YANG. Aspect sentiment analysis based on a hierarchical attention network[J]. Big Data Research, 2020, 6(5): 10.

[16] 徐志栋，陈炳阳，王晓，等.基于胶囊网络的方面级情感分类研究[J]. 智能科学与技术学报，2020，2(3)：284-292.

XU Z D, CHEN B Y, WANG X, et al. Research on capsule network-based for aspect-level sentiment classification[J]. Chinese Journal of Intelligent Science and Technology, 2020, 2(3): 284-292.

[17] 张宝华，张华平，厉铁帅，等.基于多输入模型及句法结构的中文评论情感分析方法[J]. 大数据，2021，7(6)：41-52.

ZHANG B H, ZHANG H P, LI T S, et al. Chinese comment sentiment analysis method based on multi-input model and syntactic structure[J]. Big Data Research, 2021, 7(6): 41-52.

[18] Radford A, Narasimhan K, Salimans T, et al. Improving language understanding by generative pre-training[J]. 2018.

[19] Devlin J, Chang M W, Lee K, et al. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding[J]. 2018.

[20] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]. Advances in neural information processing systems, 2017, 30.

[21] Sun C, Qiu X, Xu Y, et al. How to Fine-Tune BERT for Text Classification[J]. Springer, Cham, 2019.

[22] 杨晨，宋晓宁，宋威. SentiBERT:结合情感信息的预训练语言模型[J]. 计算机科学与探索, 2020, 14(9):8.

Chen Y, Xiaoning S, Wei S. SentiBERT: Pre-training Language Model Combining Sentiment Information[J]. Journal of Frontiers of Computer Science & Technology, 2020, 14(9): 8.