[目的/意义] 社交媒体平台上充斥着大量虚假信息、恶意链接及仇恨言论等有害内容，严重损害了用户体验。为了从根本上遏制有害内容的传播，识别并遏制发布此类信息的仇恨言论群体至关重要。[方法/过程] 从用户属性、用户生成内容、社交网络关系构建仇恨言论群体特征体系。用户属性包括性别和账号信息，用户生成内容涉及历史转发频率、情感和主题分析，社交网络采集交互特征和高影响力用户主题特征。进而引入基于二元组约束的相似度学习法，借助全连接神经网络与对比损失函数，构建仇恨言论群体发现模型。[结果/结论] 仇恨言论群体特征体系具有一定的普适性，以该特征体系为基础构建的基于相似度学习的社交媒体仇恨言论群体发现模型预测准确率达到97.38%。相较于其他的分类模型，基于相似度学习的社交媒体仇恨言论群体发现模型能够更精准地识别社交媒体中的仇恨言论群体。
[Purpose/Significance] Social media platforms are flooded with a large amount of false information, malicious links, hateful speech, and other harmful content, which seriously harms the user experience. To fundamentally curb the spread of harmful content, it is crucial to identify and suppress hate speech groups that post such information. [Method/Process] This paper constructed a feature system for hate speech groups based on user attributes, content generation, and social network relationships. User attributes included gender and account information. User-generated content involved historical retweet frequency, sentiment, and topic analysis. Social networks collected interaction features and topic features of high-influence users. And it innovatively introduced a similarity learning method based on pairwise constraints. It adopted a fully connected neural network and a contrastive loss function and constructed a discovery model for hate speech groups on social media. [Result/Conclusion] The feature system for hate speech groups proposed in this paper has a certain universality, and the prediction accuracy of the similarity learning-based discovery model based on this characteristic system reaches 97.38%. Compared with other classification models, the similarity learning-based discovery model can more accurately identify hate speech groups on social media.
[1] 中华人民共和国国务院新闻办公室. 《携手构建网络空间命 运共 同体》 白皮 书[EB/OL]. [2025-06-11]. https://www.gov.cn/xinwen/2022-11/07/content_5725220.htm. (The State Council Information Office of the People's Republic of China. "Jointly build a community with a shared future in cyberspace" white paper [EB/OL]. [2025-06-11]. https://www.gov.cn/xinwen/2022-11/07/content_5725220.htm.)
[2] 潘理. 社交网络群体行为分析[M]. 北京: 科学出版社, 2018. (PAN L. Analysis of group behavior in social networks [M]. Beijing: Science Press, 2018.)
[3] 曾子明, 王婧. 基于LDA和随机森林的微博谣言识别研究 ——以2016年雾霾谣言为例[J]. 情报学报, 2019, 38(1): 89-96. (ZENG Z M, WANG J. Research on microblog rumor identification based on LDA and random forest[J]. Journal of the China Society for Scientific and Technical Information, 2019, 38(1): 89-96.)
[4] 曾子明, 张瑜. 基于数据增强和多任务学习的突发公共卫生事件谣言识别研究[J]. 数据分析与知识发现, 2023, 7(11): 56-67. (ZENG Z M, ZHANG Y. Rumor detection of public health emergencies based on data augmentation and multi-task learning[J]. Data analysis and knowledge discovery, 2023, 7(11): 56-67.)
[5] PASCHALIDES D, STEPHANIDIS D, ANDREOU A, et al. Mandol: a big-data processing and visualization platform for monitoring and detecting online hate speech[J]. ACM transactions on Internet technology, 2020, 20(2): 1-21.
[6] WANG Z Y, WEI W, MAO X L, et al. User-based network embedding for opinion spammer detection[J]. Pattern recognition, 2022, 125: 108512.
[7] 韩忠明, 许峰敏, 段大高. 面向微博的概率图水军识别模型[J]. 计算机研究与发展, 2013, 50(S2): 180-186. (HAN Z M, XU F M, DUAN D G. Probabilistic graphical model for identifying water army in microblogging system[J]. Journal of computer research and development, 2013, 50(S2): 180-186.)
[8] YING H, PIN Y, PENG S C. Semi-supervised Internet water army detection based on graph embedding[J]. Multimedia tools and applications, 2022, 82(7): 9891-9912.
[9] 王法强, 张宏志, 王鹏, 等. 计算机视觉中相似度学习方法的研究进展[J]. 智能计算机与应用, 2019, 9(1): 149-152, 158. (WANG F Q, ZHANG H Z, WANG P, et al. Research progress on the similarity learning methods in computer vision[J]. Intelligent computer and applications, 2019, 9(1): 149-152, 158.)
[10] ZHANG R M, LIN L, ZHANG R, et al. Bit-scalable deep hashing with regularized similarity learning for image retrieval and person re-identification[J]. IEEE transactions on image processing : a publication of the IEEE signal processing society, 2015, 24(12): 4766-79.
[11] GUILLAUMIN M, VERBEEK J, SCHMID C. Is that you? Metric learning approaches for face identification[C]//The 12th IEEE international conference on computer vision. Kyoto: IEEE, 2009: 498-505.
[12] ZHENG W S, GONG S G, XIANG T. Person reidentification by probabilistic relative distance comparison[C]//The 24th IEEE conference on computer vision and pattern recognition. Colorado: IEEE, 2011: 649-656.
[13] SUN Y, CHEN Y H, WANG X G, et al. Deep learning face representation by joint identification-verification[C]//NIPS'14 proceedings of the 27th international conference on neural information processing systems-volume 2. Montreal: ACM, 2014, 27: 1988-1996.
[14] 刘蓉, 陈波, 于泠, 等. 恶意社交机器人检测技术研究[J]. 通信学报, 2017, 38(S2): 197-210. (LIU R, CHEN B, YU L, et al. Overview of detection techniques for malicious social bots[J]. Journal on communications, 2017, 38(S2): 197-210.)
[15] 汪子航, 言鹏韦, 蒋卓人. 基于可解释图神经网络模型的社交媒体谣言识别研究[J]. 情报学报, 2023, 42(11): 1369-1381. (WANG Z H, YAN P W, JIANG Z R. Interpretable graph neural network for social media rumor detection[J]. Journal of the China Society for Scientific and Technical Information, 2023, 42(11): 1369-1381.)
[16] 夏崇欢, 李华康, 孙国梓. 基于行为特征分析的微博恶意用户识别[J].计算机科学, 2018, 45(12): 111-116. (XIA C H, LI H K, SUN G Z. Microblogging malicious user identification based on behavior characteristic analysis[J]. Journal of computer science and technology, 2018, 45(12): 111-116.)
[17] STEVENS M E. Introduction to the special issue on optical character recognition (OCR)[J]. Pattern recognition, 1970, 2(3): 147-150.
[18] CHEN H L, SHI L, WU P H, et al. Fine-grained sentiment analysis of Chinese reviews using LSTM network[J]. Journal of engineering science and technology review, 2018, 11(1): 174- 179.
[19] 微博情感数据集[EB/OL]. [2025-03-14]. https://pan.baidu.com/s/1wlorPLGsyjEMJgx8Noxx-A?pwd=d0rg&_at_=1728439597541 Weibo emotion dataset [EB/OL]. [2025-03-14]. https://pan. baidu.com/s/1wlorPLGsyjEMJgx8Noxx-A?pwd=d0rg&_ at_=1728439597541.
[20] BLEI D M, NG A Y, JORDAN M I. Latent dirichlet allocation[J]. Journal of machine learning research, 2003, 3(1): 993-1022.
[21] BREIMAN L. Random forests[J]. Mach learn, 2001, 45(1): 5-32.
[22] HADSELL R, CHOPRA S, LECUN Y. Dimensionality reduction by learning an invariant mapping[C]//2006 IEEE computer society conference on computer vision and pattern recognition. New York: IEEE, 2006, 2: 1735-1742.
[23] COVER T M, HART P E. Nearest neighbor pattern classification[J]. IEEE transactions on information theory, 1968, 13: 21-27.
[24] KINGMA D, BA J. Adam: a method for stochastic optimization[EB/OL]. [2025-03-14]. https://arxiv.org/abs/1412.6980.
