1. Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, China; 2. Yunnan Key Laboratory of Artificial Intelligence, Kunming University of Science and Technology, Kunming 650500, China
Abstract：[Objective] Social event detection involves the identification of trending events from various social event contents. It has garnered widespread attention in recent years. However, in practical scenarios, the performance of social event detection suffers in the case of data scarcity. Moreover, privacy concerns have resulted in regulatory restrictions preventing organizations from sharing data without explicit user consent, which makes centralized data training impractical. The current approaches primarily address the issue of data scarcity through cross-lingual knowledge transfer but often ignore the challenges associated with data privacy. Consequently, this study proposes a framework for cross-lingual social event detection based on federated knowledge distillation, referred to as FedEvent, with the goal of distilling knowledge from high-resource clients to low-resource ones.[Methods] The framework employs parameter-efficient fine-tuning techniques and triple contrastive losses to effectively map non-English semantic spaces to English ones, and employs a federated distillation strategy to ensure data privacy. In addition, a four-stage lifecycle mechanism is designed to adapt to incremental scenarios. It includes four stages: pre-training of high-resource clients, pre-training of low-resource clients, client detection, and maintenance. In the first stage, a specific algorithm is used to train the initial model. In the second stage, the model trained in the first stage uses the central server as a medium for knowledge transfer to assist low-resource clients in training the initial model. In the third stage, the trained model is used to directly detect each incoming message. In the fourth stage, the model is continuously trained with the latest message blocks, and a federated codistillation mechanism is used to enable online learning for each client, allowing for the model to learn new knowledge. Based on the large-scale public English dataset Events2012, this study supplements event samples in Chinese and Vietnamese according to its topic descriptions, constructs private datasets for low-resource clients (Chinese-language client and Vietnamese-language client), and thereby establishes a multi-client, cross-lingual experimental environment. The performance of the framework is evaluated on the aforementioned datasets across two widely used clustering evaluation metrics: Normalized Mutual Information (NMI) and Adjusted Mutual Information (AMI).[Results] 1) The experimental results demonstrate that compared with the existing methods, the proposed framework achieves effective knowledge transfer while ensuring data privacy. 2) In comparison with the single-node setup, the proposed framework demonstrates notable enhancement in the multimode environment on the Chinese-language client from 1.6% to 204.3% in NMI and from 2.0% to 342.9% in AMI. On the Vietnamese-language client, the respective improvements are noted to be between 2.3% to 6 200% and 0 to 4 400%. 3) The proposed framework performs very similarly to the state-of-the-art centralized method Cross-Lingual Knowledge Distillation (CLKD). 4) Case analysis reveals that FedEvent's outstanding performance on clusters with high-resource clients significantly influences its effectiveness on analogous clusters with low-resource clients. This demonstrates that FedEvent can effectively transfer knowledge from high-resource clients to low-resource ones. Furthermore, visual analysis vividly highlights the superior clustering outcomes achieved by FedEvent.[Conclusions] The framework employs a lifecycle mechanism to accommodate the needs of event detection both in online and offline scenarios, effectively transferring knowledge through process and outcome supervision. Using knowledge distillation techniques, it mitigates the challenges faced when addressing low-resource languages and leverages a cross-lingual word embedding module to map semantic spaces between various languages. The proposed framework achieved the expected results, notably enhancing the model's performance.
[1] ATEFEH F, KHREICH W. A survey of techniques for event detection in twitter[J]. Computational Intelligence, 2015, 31(1):132-164. [2] PENG H, LI J X, SONG Y Q, et al. Streaming social event detection and evolution discovery in heterogeneous information networks[J]. ACM Transactions on Knowledge Discovery from Data, 2021, 15(5):89. [3] GASPAR R, PEDRO C, PANAGIOTOPOULOS P, et al. Beyond positive or negative:Qualitative sentiment analysis of social media reactions to unexpected stressful events[J]. Computers in Human Behavior, 2016, 56:179-191. [4] NISAR T M, YEUNG M. Twitter as a tool for forecasting stock market movements:A short-window event study[J]. The journal of finance and data science, 2018, 4(2):101-119. [5] MAROZZO F, BESSI A. Analyzing polarization of social media users and news sites during political campaigns[J]. Social Network Analysis and Mining, 2018, 8(1):1. [6] AGGARWAL C C, SUBBIAN K. Event detection in social Streams[C]//Proceedings of the 12th SIAM International Conference on Data Mining. Anaheim, USA:Society for Industrial and Applied Mathematics, 2012:624-635. [7] HU L M, ZHANG B, HOU L, et al. Adaptive online event detection in news streams[J]. Knowledge-Based Systems, 2017, 138:105-112. [8] OZDIKIS O, KARAGOZ P, O AG UZTÜZÜN H. Incremental clustering with vector expansion for online event detection in microblogs[J]. Social Network Analysis and Mining, 2017, 7(1):56. [9] ZHANG K, ZI J, WU L G. New event detection based on indexing-tree and named entity[C]//Proceedings of the 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. Amsterdam, The Netherlands, USA:Association for Computing Machinery 2007:215-222. [10] FEDORYSZAK M, FREDERICK B, RAJARAM V, et al. Real-time event detection on social data streams[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery&Data Mining. Anchorage, USA:Association for Computing Machinery 2019:2774-2782. [11] LIU B, HAN F X, NIU D, et al. Story forest:Extracting events and telling stories from breaking news[J]. ACM Transactions on Knowledge Discovery from Data, 2020, 14(3):31. [12] LIU F Z, XUE S, WU J, et al. Deep learning for community detection:progress, challenges and opportunities[C]//Proceedings of the 29th International Joint Conference on Artificial Intelligence. Yokohama, Japan:ijcai.org, 2021:4981-4987. [13] LIU Y P, PENG H, LI J X, et al. Event detection and evolution in multi-lingual social streams[J]. Frontiers of Computer Science, 2020, 14(5):145612. [14] YU W R, LI J X, BHUIYAN M Z A, et al. Ring:Real-time emerging anomaly monitoring system over text streams[J]. IEEE Transactions on Big Data, 2019, 5(4):506-519. [15] BECKER H, NAAMAN M, GRAVANO L. Beyond trending topics:Real-world event identification on twitter[C]//Proceedings of the 5th International AAAI Conference on Web and Social Media. Barcelona, Spain:AAAI Press, 2021:438-441. [16] CHENG X Q, YAN X H, LAN Y Y, et al. Btm:Topic modeling over short texts[J]. IEEE Transactions on Knowledge and Data Engineering, 2014, 26(12):2928-2941. [17] CORDEIRO M. Twitter event detection:Combining wavelet analysis and topic inference summarization[C]//Proceedings of the Doctoral Symposium on Informatics Engineering. Porto, Portugal:University of Porto, 2012:11-16. [18] ZHOU X M, CHEN L. Event detection over twitter social media streams[J]. The VLDB journal, 2014, 23(3):381-400. [19] Chung H W, Garrette D, Tan K C, et al. Improving multilingual models with language-clustered vocabularies[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). Pennsylvania, USA:Association for Computational Linguistics, 2020:4536-4546. [20] SU X, XUE S, LIU F Z, et al. A comprehensive survey on community detection with deep learning[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35(4):4682-4702. [21] CUI W Q, DU J P, WANG D W, et al. MVGAN:Multi-view graph attention network for social event detection[J]. ACM Transactions on Intelligent Systems and Technology, 2021, 12(3):27. [22] PENG H, LI J X, GONG Q R, et al. Fine-grained event categorization with heterogeneous graph convolutional networks[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence. Macao China:ijcai.org, 2019:3238-3245. [23] PENG H, ZHANG R T, LI S N, et al. Reinforced, incremental and cross-lingual event detection from social messages[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(1):980-998. [24] REN J Q, JIANG L, PENG H, et al. From known to unknown:Quality-aware self-improving graph neural network for open set social event detection[C]//Proceedings of the 31st ACM International Conference on Information&Knowledge Management. Atlanta, USA:Association for Computing Machinery, 2022:1696-1705. [25] REN J Q, JIANG L, PENG H, et al. Evidential temporal-aware graph-based social event detection via dempster-shafer theory[C]//2022 IEEE International Conference on Web Services (ICWS). Barcelona, Spain:IEEE Computer Society Press, 2022:331-336. [26] CAO Y W, PENG H, WU J, et al. Knowledge-preserving incremental social event detection via heterogeneous gnns[C]//Proceedings of the Web Conference 2021. Ljubljana, Slovenia:Association for Computing Machinery, 2021:3383-3395. [27] MOHIUDDIN T, BARI M S, JOTY S. LNMap:Departures from isomorphic assumption in bilingual lexicon induction through non-linear maping in latent space[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). Pennsylvania, USA:Association for Computational Linguistics, 2020:2712-2723. [28] REN J Q, PENG H, JIANG L, et al. Transferring knowledge distillation for multilingual social event detection[EB/OL].(2021-08-13)[2024-08-24] . https://doi.org/10.03084/arXiv.210 8.03084. [29] HU E J, SHEN Y L, WALLIS P, et al. Lora:Low-rank adaptation of large language models[C]//Tenth International Conference On learning Representations. Ithaca, NY:Springer, 2022. [30] CAMPELLO R J G B, MOULAVI D, SANDER J. Density-based clustering based on hierarchical density estimates[C]//17th Pacific-Asia Conference on Knowledge Discovery and Data Mining. Gold Coast, Australia:Springer, 2013:160-172. [31] WU C H, WU F Z, LYU L J, et al. Communication-efficient federated learning via knowledge distillation[J]. Nature Communications, 2022, 13(1):2032. [32] PIRES T, SCHLINGER E, GARRETTE D. How multilingual is multilingual BERT?[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Florence, Italy:Association for Computational Linguistics, 2019:4996-5001. [33] MCMAHAN B, MOORE E, RAMAGE D, et al. Communication-efficient learning of deep networks from decentralized data[C]//Proceedings of the 20th International Conference on Artificial Intelligence and Statistics. Fort Lauderdale, USA:PMLR, 2017:1273-1282. [34] Li P, YU X Y, PENG H, et al. Relational prompt-based pre-trained language models for social event detection[J]. ACM Transactions on Information Systems, 2024, 1046-8188.DOI:10.1145/3695869. [35] VAN DER, MAATEN L, HINTON G. Visualizing data using t-SNE[J]. Journal of machine learning research, 2008, 9(86):2579-2605.
2025, Vol. 65 
