计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (3): 118-127.doi: 10.11896/jsjkx.221200054

• 数据库&大数据&数据科学 • 上一篇    下一篇

基于局部数据增强动态图的事件预测

潘磊1, 刘欣2, 陈君益2, 程章桃2, 刘乐源2, 周帆2,3   

  1. 1 中国电子科技集团公司第十研究所 成都610036
    2 电子科技大学信息与软件工程学院 成都610054
    3 喀什地区电子信息产业技术研究院 新疆 喀什844099
  • 收稿日期:2022-12-08 修回日期:2023-04-07 出版日期:2024-03-15 发布日期:2024-03-13
  • 通讯作者: 刘乐源(leyuanliu@uestc.edu.cn)
  • 作者简介:(mapan.lei@163.com)
  • 基金资助:
    国家自然科学基金(62176043,62072077);四川省自然科学基金(2022NSFC0505);四川省科技计划(2022YFSY0006);厅市共建智能终端四川省重点实验室开放课题(SCITLAB-20006)

Event Prediction Based on Dynamic Graph with Local Data Augmentation

PAN Lei1, LIU Xin2, CHEN Junyi2, CHENG Zhangtao2, LIU Leyuan2, ZHOU Fan2,3   

  1. 1 No.10 Research Institute of China Electronics Technology Group Corporation,Chengdu 610036,China
    2 School of Information and Software Engineering,University of Electronic Science and Technology of China,Chengdu 610054,China
    3 Information Industry Technology Research Institute of Kashi Region,Kashi,Xinjiang 844099,China
  • Received:2022-12-08 Revised:2023-04-07 Online:2024-03-15 Published:2024-03-13
  • About author:PAN Lei,born in 1986,Ph.D,senior engineer.His main research interests include NLP,multimodal data feature extraction,crisis event analysis and intelligent text generation.LIU Leyuan,born in 1982,Ph.D,research associate.His main research interests include graph learning,social network data mining and event prediction.
  • Supported by:
    National Natural Science Foundation of China(62176043,62072077),Natural Science Foundation of Sichuan Province,China(2022NSFC0505),Sichuan Science and Technology Program(2022YFSY0006) and Open Project of Intelligent Terminal Key Laboratory of Sichuan Province(SCITLAB-20006).

PDF (PC)

94

摘要： 事件指在真实世界中特定的时间和地点发生的与特定主题相关的活动,例如,社会动乱、暴恐袭击、自然灾害和传染病流行等事件会对国家安全和人民群众的生活产生重大威胁。如果能对此类事件的发生进行有效预测,将最大程度地减少负面事件带来的影响或最大化正面事件带来的利益。关于事件的研究中,准确预测事件仍然是一个非常具有挑战性的任务。文中提出了一种基于图注意力网络的事件预测方法LAT-GAT(Local Augmented Temporal-GAT),该方法使用条件变分编码器,在所构建的事件图中对目标节点的邻居节点生成新的特征样本,与节点原有特征进行拼合,形成新的节点特征,实现了对事件的传播结构的利用;另外,LAT-GAT还考虑了历史事件发生的时间先后顺序,将网络在上一时间点的输出结果集成到当前时间的特征中,从而实现了对事件传播时间特性的利用。最后,在泰国、印度、埃及和俄罗斯这4个国家真实事件数据集上,与多种代表性基线方法进行了对比实验。实验结果表明,LAT-GAT在4个国家数据上的F1评分都优于基线方法;在泰国、俄罗斯和印度数据集上召回率优于基线方法;在泰国、埃及和印度数据集上也获得了最高的准确率。还通过消融实验考察了模型参数对最终结果的影响。

关键词: 事件预测, 图注意力网络, 动态图, 条件变分编码器, 数据增强

Abstract: Event refers to activities that occur in real world at specific time and places.For instance,unrest,violent terrorist attacks,natural disasters and the spread of infectious diseases,will bring great threats and losses to national security and human life.If the occurrence of such events could be predicted more precisely and effectively,the impact of negative events will be minimized,and it is possible to maximize the benefits of the positive events.It is still a very challenging task to predict events accurately.An event prediction method named local augmented temporal-GAT(LAT-GAT) based on graph attention network is proposed in this paper.It uses conditional variational encoders to generate new features,which will be concatenated with the original features to new one,based on neighbors of the current node.With this approach,our model can utilize the propagation structure of events.In addition,the chronological order of events occurrence is considered by our model.The feature of events in last time point is integrated into the output of the neural network in current time.The temporal property of event propagation is exploited through temporal data integration.And finally,the proposed method is compared with a number of representative baseline me-thods on the real-world datasets,including Thailand,India,Egypt and Russia.The results show that LAT-GAT has the best F1 scores in all datasets.The recall of our model exceeds that of any other baseline methods in the datasets of Thailand,Russia and India.In Thailand,Egypt and India,our model achieves the best precision.Ablation experiments are also conducted to investigate the influence of the model parameters on the final results.

Key words: Event prediction, Graph attention network, Dynamic graph, Conditional variational auto-encoder, Data augmentation

中图分类号: 

  • TP391
[1]ZHAO L.Event prediction in the big data era:A systematic survey[J].ACM Computing Surveys(CSUR),2021,54(5):1-37.
[2]ZHAO L,YE J,CHEN F,et al.Hierarchical incomplete multi-source feature learning for spatiotemporal event forecasting[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.2016:2085-2094.
[3]INCEOGLU F,JEPPESEN J H,KONGSTAD P,et al.Usingmachine learning methods to forecast if solar flares will be associated with CMEs and SEPs[J].The Astrophysical Journal,2018,861(2):128.
[4]DE CAIGNY A,COUSSEMENT K,DE BOCK K W.A new hybrid classification algorithm for customer churn prediction based on logistic regression and decision trees[J].European Journal of Operational Research,2018,269(2):760-772.
[5]QIAO Z,ZHAO S,XIAO C,et al.Pairwise-ranking based collaborative recurrent neural networks for clinical event prediction[C]//Proceedings of the Twenty-seventh International Joint Conference on Artificial Intelligence.2018.
[6]SIMMA A,JORDAN M I.Modeling events with cascades ofPoisson processes[J].arXiv:1203.3516,2012.
[7]HOU X L,ZHOU P P,ZHAO J B.An automatic exposure mo-del of image sequence acquisition for HDR scenes[J].Journal of Chongqing University of Technology:Natural Science,2022,36(4):153-161.
[8]BERHICH A,BELOUADHA F Z,KABBAJ M I.An attention-based LSTM network for large earthquake prediction[J].Soil Dynamics and Earthquake Engineering,2023,165:107663.
[9]RAMA-MANEIRO E,VIDAL J C,LAMA M.Embedding graph convolutional networks in recurrent neural networks for predictive monitoring[J].IEEE Transactions on Knowledge and Data Engineering,2024,36(1):137-151.
[10]HAO M,JIANG D,DING F,et al.Simulating spatio-temporal patterns of terrorism incidents on the Indochina Peninsula with GIS and the random forest method[J].ISPRS International Journal of Geo-Information,2019,8(3):133.
[11]PIRAJAN F,FAJARDO A,MELGAREJO M.Towards a deep learning approach for urban crime forecasting[C]//Workshop on Engineering Applications.Cham:Springer,2019:179-189.
[12]HAN J,PEI J,TONG H.Data mining:concepts and techniques[M].Morgan kaufmann,2022.
[13]SU Y T,WANG J,ZHAO W,et al.Dynamic graph convolu-tional neural network for image sentiment distribution prediction[J].Journal of Jilin University(Engineering and Technology Edition),2023,53(9):2601-2610.
[14]RADINSKY K,DAVIDOVICH S,MARKOVITCH S.Learning causality for news events prediction[C]//Proceedings of the 21st International Conference on World Wide Web.2012:909-918.
[15]LEI L,REN X,FRANCISCUS N,et al.Event prediction based on causality reasoning[C]//Asian Conference on Intelligent Information and Database Systems.Cham:Springer,2019:165-176.
[16]YANG Y,WEI Z,CHEN Q,et al.Using external knowledge for financial event prediction based on graph neural networks[C]//Proceedings of the 28th ACM International Conference on Information and Knowledge Management.2019:2161-2164.
[17]LIU S,YING R,DONG H,et al.Local augmentation for graph neural networks[C]//International Conference on Machine Learning.PMLR,2022:14054-14072.
[18]LI Q,HAN Z,WU X M.Deeper insights into graph convolutional networks for semi-supervised learning[C]//Thirty-Se-cond AAAI Conference on Artificial Intelligence.2018.
[19]SHU K,SLIVA A,WANG S,et al.Fake news detection on social media:A data mining perspective[J].ACM SIGKDD Explorations Newsletter,2017,19(1):22-36.
[20]DENG S,RANGWALA H,NING Y.Learning dynamic context graphs for predicting social events[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Disco-very & Data Mining.2019:1007-1016.
[21]BILOŠ M,CHARPENTIER B,GÜNNEMANN S.Uncertainty on asynchronous time event prediction[J].arXiv:1911.05503,2019.
[22]WANG Q,JIN G,ZHAO X,et al.CSAN:A neural networkbenchmark model for crime forecasting in spatio-temporal scale[J].Knowledge-Based Systems,2020,189:105120.
[23]YUAN Z,ZHOU X,YANG T.Hetero-convlstm:A deep lear-ning approach to traffic accident prediction on heterogeneous spatio-temporal data[C]//Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.2018:984-992.
[24]DENG S,RANGWALA H,NING Y.Dynamic knowledge graph based multi-event forecasting[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Disco-very & Data Mining.2020:1585-1595.
[25]DU L,DING X,ZHANG Y,et al.A Graph Enhanced BERTModel for Event Prediction[C]//Findings of the Association for Computational Linguistics:ACL 2022.2022:2628-2638.
[26]LIU X,ZHANG F,HOU Z,et al.Self-supervised learning:Ge-nerative or contrastive[J].arXiv:2006.08218,2021.
[27]LI X,LIU X P,LI W C,et al.Survey on Contrastive Learning Research[J].Journal of Chinese Computer Systems,2023,44(4):787-797.
[28]HE K,FAN H,WU Y,et al.Momentum contrast for unsuper-vised visual representation learning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:9729-9738.
[29]CHEN T,KORNBLITH S,NOROUZI M,et al.A simpleframework for contrastive learning of visual representations[C]//International Conference on Machine Learning.PMLR,2020:1597-1607.
[30]TIAN Y,CHEN X,GANGULI S.Understanding self-super-vised learning dynamics without contrastive pairs[C]//International Conference on Machine Learning.PMLR,2021:10268-10278.
[31]CHEN X,HE K.Exploring simple siamese representation lear-ning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:15750-15758.
[32]GOODFELLOW I,POUGET-ABADIE J,MIRZA M,et al.Ge-nerative adversarial networks[J].Communications of the ACM,2020,63(11):139-144.
[33]BOWMAN S R,VILNIS L,VINYALS O,et al.Generating sentences from a continuous space[C]//20th SIGNLL Conference on Computational Natural Language Learning,CoNLL 2016.Association for Computational Linguistics(ACL),2016:10-21.
[34]YOU J,YING R,REN X,et al.Graphrnn:Generating realistic graphs with deep auto-regressive models[C]//International Conference on Machine Learning.PMLR,2018:5708-5717.
[35]YOU J,LIU B,YING Z,et al.Graph convolutional policy network for goal-directed molecular graph generation[J].arXiv:1806.02473,2018.
[36]HU Z,DONG Y,WANG K,et al.Gpt-gnn:Generative pre-training of graph neural networks[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Disco-very & Data Mining.2020:1857-1867.
[37]WANG C,PAN S,LONG G,et al.Mgae:Marginalized graphautoencoder for graph clustering[C]//Proceedings of the 2017 ACM on Conference on Information and Knowledge Management.2017:889-898.
[38]SALEHI A,DAVULCU H.Graph Attention Auto-Encoders[C]//2020 IEEE 32nd International Conference on Tools with Artificial Intelligence(ICTAI).IEEE Computer Society,2020:989-996.
[39]TANG M,YANG C,LI P.Graph Auto-Encoder via Neighborhood Wasserstein Reconstruction[J].arXiv:2202.09025,2022.
[40]HOU Z Y,LIU X,CEN Y K,et al.GraphMAE:Self-Super-vised Masked Graph Autoencoders[C]//The 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining(KDD'22).2022:594-604.
[41]MIKOLOV T,SUTSKEVER I,CHEN K,et al.Distributed representations of words and phrases and their compositionality[C]//Proceedings of the 26th International Conference on Neural Information Processing Systems.2013:3111-3119.
[42]CHURCH K,HANKS P.Word association norms,mutual information,and lexicography[J].Computational Linguistics,1990,16(1):22-29.
[43]KINGMA D P,WELLING M.Auto-Encoding Variational Bayes[J].arXiv:1312.6114,2014.
[44]SOHN K,LEE H,YAN X.Learning structured output repre-sentation using deep conditional generative models[C]//Proceedings of the 28th International Conference on Neural Information Processing Systems.2015:3483-3491.
[45]ELIZABETH B,JENNIFER L,SEAN O,et al.ICEWS Coded Event Data.Harvard Dataverse[EB/OL].[2023-12-03].https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/28075.
[46]PAREJA A,DOMENICONI G,CHEN J,et al.Evolvegcn:Evolving graph convolutional networks for dynamic graphs[J].ar-Xiv:1902.10191,2020.
[47]HU Z N,DONG Y X,WANG K S,et al.Heterogeneous graph transformer[C]//Proceedings of The Web Conference 2020.2020:2704-2710.
[48]WU Z H,PAN S R,LONG G D,et al.Graph wavenet for deep spatial-temporal graph modeling[J].arXiv:1906.00121,2019.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发