Personal event detection method based on text mining in social media

doi:10.11772/j.issn.1001-9081.2022010106

Journal of Computer Applications ›› 2022, Vol. 42 ›› Issue (11): 3513-3519.DOI: 10.11772/j.issn.1001-9081.2022010106

• ChinaService 2021 • Previous Articles

Personal event detection method based on text mining in social media

Rui XIAO, Mingyi LIU, Zhiying TU, Zhongjie WANG()

Faculty of Computing，Harbin Institute of Technology，Harbin Heilongjiang 150001，China

Received:2022-01-27 Revised:2022-03-20 Accepted:2022-04-02 Online:2022-11-14 Published:2022-11-10
Contact: Zhongjie WANG
About author:XIAO Rui， born in 1997， M. S. candidate. His research interests include service computing.
LIU Mingyi， born in 1995， Ph. D. His research interests include service computing.
TU Zhiying， born in 1983， Ph. D.， associate professor. His research interests include software engineering， service computing， knowledge engineering， enterprise business modeling.
WANG Zhongjie， born in 1978， Ph. D.， professor. His research interests include service computing， software engineering.
Supported by:
National Natural Science Foundation of China(61772155)

基于社交媒体文本挖掘的个人事件检测方法

肖锐, 刘明义, 涂志莹, 王忠杰()

哈尔滨工业大学计算学部，哈尔滨 150001

通讯作者: 王忠杰
作者简介:肖锐（1997—），男，重庆人，硕士研究生，主要研究方向：服务计算
刘明义（1995—），男，安徽宣城人，博士，CCF会员，主要研究方向：服务计算
涂志莹（1983—），男，福建龙岩人，副教授，博士，CCF会员，主要研究方向：软件工程、服务计算、知识工程、企业业务建模
王忠杰（1978—），男，山东龙口人，教授，博士，CCF高级会员，主要研究方向：服务计算、软件工程。rainy@hit.edu.cn
基金资助:
国家自然科学基金资助项目(61772155)

Abstract

Abstract:

Users’ social media contains their past personal experiences and potential life patterns， and the study of their patterns is of great value for predicting users’ future behaviors and performing personalized recommendations for users. By collecting Weibo data， 11 types of events were defined， and a three?stage Pipeline system was proposed to detect personal events by using BERT （Bidirectional Encoder Representations from Transformers） pre?trained models in three stages respectively， including BERT+BiLSTM+Attention， BERT+FullConnect and BERT+BiLSTM+CRF. The information of whether the text contained defined events， the event types of events contained， and the elements contained in each event were extracted from the Weibo， and the specific elements are Subject （subject of the event）， Object （event element）， Time （event occurrence time）， Place （place where the event occurred） and Tense （tense of the event）， thereby exploring the change law of user’s personal event timeline to predict personal events. Comparative experiments and analysis were conducted with classification algorithms such as logistic regression， naive Bayes， random forest and decision tree on a collected real user Weibo dataset. Experimental results show that the BERT+BiLSTM+Attention， BERT+FullConnect， BERT+BiLSTM+CRF methods used in three stages achieve the highest F1?score， verifying the effectiveness of the proposed methods. Finally， the personal event timeline was visually built according to the extracted events with time information.

Key words: social media, personal event, event detection, BERT (Bidirectional Encoder Representations from Transformers) model, personal event timeline

摘要：

用户的社交媒体中蕴含着他们过去的个人经历和潜在的生活规律，研究其规律对预测用户未来的行为以及对用户进行个性化推荐有很大的价值。通过收集微博数据，定义了11种类型的事件，并提出了一个三阶段的Pipeline的系统，利用BERT预训练模型，分别在三个阶段使用BERT+BiLSTM+Attention、BERT+FullConnect、BERT+BiLSTM+CRF方法进行个人事件检测。从微博文本中抽取出该文本是否包含定义的事件、包含的事件类型、每种事件包含的元素等信息，具体元素为Subject（事件主语）、Object（事件元素）、Time（事件发生时间）、Place（事件发生的地点）和Tense（事件发生的时态），从而探究用户个人时间轴上的事件变化规律来预测个人事件。在收集的真实用户微博数据集上进行实验，并与逻辑回归、朴素贝叶斯、随机森林、决策树等分类算法进行对比分析。实验结果表明，三个阶段中的BERT+BiLSTM+Attention、BERT+FullConnect和BERT+BiLSTM+CRF方法均取得了最高的F1值，验证了所提方法的有效性。最后根据所提方法抽取出的事件和其中的时间信息可视化地构建了用户的个人事件时间轴。

关键词: 社交媒体, 个人事件, 事件检测, BERT模型, 个人事件时间轴

CLC Number:

TP391.1

Rui XIAO, Mingyi LIU, Zhiying TU, Zhongjie WANG. Personal event detection method based on text mining in social media[J]. Journal of Computer Applications, 2022, 42(11): 3513-3519.

肖锐, 刘明义, 涂志莹, 王忠杰. 基于社交媒体文本挖掘的个人事件检测方法[J]. 《计算机应用》唯一官方网站, 2022, 42(11): 3513-3519.

Figures/Tables 13

Fig. 1 Example of personal event timeline and recommendations of prediction

Tab. 1 Examples of event extraction from Weibo text

微博

是否包含

定义的事件

（audiovisual）‒（User［用户］，盲山，昨天，‒，过去）

老爸生日，我们今天吃到了各种菌类：羊肚菌、黑松露

是

birth，eat

（birth）‒（老爸，生日，‒，‒，现在）

（eat）‒（我们，羊肚菌、黑松露，今天，‒，现在）

Fig. 2 Personal event extraction framework

Fig. 3 BERT+BiLSTM+Attention classification model

Fig. 4 BERT+FullConnect classification model

Fig. 5 BERT+BiLSTM+CRF classification model

Fig. 6 Example of part of data in dataset

Tab. 2 Personal event detection results

模型	$A$	$P$	$R$	$F 1$
logistic regression	75.16	73.86	56.32	63.91
naive Bayes	67.83	63.45	65.33	64.38
BERT+FullConnect	86.53	80.25	80.74	80.49
BERT+BiLSTM	87.51	83.01	80.43	81.70
BERT+BiLSTM+Attention	88.63	80.18	85.42	82.71

Tab. 2 Personal event detection results

模型	$A$	$P$	$R$	$F 1$
logistic regression	75.16	73.86	56.32	63.91
naive Bayes	67.83	63.45	65.33	64.38
BERT+FullConnect	86.53	80.25	80.74	80.49
BERT+BiLSTM	87.51	83.01	80.43	81.70
BERT+BiLSTM+Attention	88.63	80.18	85.42	82.71

Tab. 3 Personal event multi?label classification results

模型	$A$	$P$	$F 1$
KNN	86.21	46.34	60.28
random forest	83.46	48.72	61.52
decision tree	73.10	61.63	66.59
BERT+BiLSTM	88.05	88.81	88.08
BERT+BiLSTM+Attention	86.54	90.26	88.09
BERT+FullConnect	91.18	92.88	91.67

Tab. 3 Personal event multi?label classification results

模型	$A$	$P$	$F 1$
KNN	86.21	46.34	60.28
random forest	83.46	48.72	61.52
decision tree	73.10	61.63	66.59
BERT+BiLSTM	88.05	88.81	88.08
BERT+BiLSTM+Attention	86.54	90.26	88.09
BERT+FullConnect	91.18	92.88	91.67

Tab. 4 Personal event element extraction results on word level

模型	$A$	$P$	$F 1$
CRF	52.72	15.65	24.13
BiLSTM+CRF	67.83	47.67	55.99
BERT+FullConnect	77.13	73.23	74.90
BERT+CRF	78.60	75.50	76.92
BERT+BiLSTM+CRF	80.41	74.75	77.38

Tab. 4 Personal event element extraction results on word level

模型	$A$	$P$	$F 1$
CRF	52.72	15.65	24.13
BiLSTM+CRF	67.83	47.67	55.99
BERT+FullConnect	77.13	73.23	74.90
BERT+CRF	78.60	75.50	76.92
BERT+BiLSTM+CRF	80.41	74.75	77.38

Tab. 5 Personal event elements extraction results on entity level

模型	$F 1$
模型	Overall	Sub	Obj	Time	Place
CRF	18.57	21.88	13.63	31.32	8.62
BiLSTM+CRF	52.81	49.33	53.94	55.12	47.24
BERT+FullConnect	68.12	64.43	70.05	62.34	66.08
BERT+CRF	68.44	66.03	70.92	62.52	68.43
BERT+BiLSTM+CRF	69.46	66.74	70.76	65.77	70.53

Tab. 5 Personal event elements extraction results on entity level

模型	$F 1$
模型	Overall	Sub	Obj	Time	Place
CRF	18.57	21.88	13.63	31.32	8.62
BiLSTM+CRF	52.81	49.33	53.94	55.12	47.24
BERT+FullConnect	68.12	64.43	70.05	62.34	66.08
BERT+CRF	68.44	66.03	70.92	62.52	68.43
BERT+BiLSTM+CRF	69.46	66.74	70.76	65.77	70.53

Tab. 6 Extraction results of whole Pipeline system

阶段1模型	$F 1$	阶段2模型	$F 1$	阶段3模型	$F 1$
BERT+BiLSTM	82.05	BERT+BiLSTM+Attention	79.61	BERT+CRF	69.82
		BERT+BiLSTM+Attention	79.61	BERT+BiLSTM+CRF	71.87
		BERT+FullConnect	81.12	BERT+CRF	70.61
		BERT+FullConnect	81.12	BERT+BiLSTM+CRF	72.49
BERT+BiLSTM+Attention	83.72	BERT+BiLSTM+Attention	80.08	BERT+CRF	70.30
		BERT+BiLSTM+Attention	80.08	BERT+BiLSTM+CRF	72.13
		BERT+FullConnect	82.78	BERT+CRF	70.43
		BERT+FullConnect	82.78	BERT+BiLSTM+CRF	73.41

Tab. 6 Extraction results of whole Pipeline system

阶段1模型	$F 1$	阶段2模型	$F 1$	阶段3模型	$F 1$
BERT+BiLSTM	82.05	BERT+BiLSTM+Attention	79.61	BERT+CRF	69.82
		BERT+BiLSTM+Attention	79.61	BERT+BiLSTM+CRF	71.87
		BERT+FullConnect	81.12	BERT+CRF	70.61
		BERT+FullConnect	81.12	BERT+BiLSTM+CRF	72.49
BERT+BiLSTM+Attention	83.72	BERT+BiLSTM+Attention	80.08	BERT+CRF	70.30
		BERT+BiLSTM+Attention	80.08	BERT+BiLSTM+CRF	72.13
		BERT+FullConnect	82.78	BERT+CRF	70.43
		BERT+FullConnect	82.78	BERT+BiLSTM+CRF	73.41

Fig. 7 Event timeline display

References 18

1	DI EUGENIO B， GREEN N， SUBBA R. Detecting life events in feeds from Twitter［C］// Proceedings of the IEEE 7th International Conference on Semantic Computing. Piscataway： IEEE， 2013： 274-277. 10.1109/icsc.2013.54
2	DICKINSON T， FERNÁNDEZ M， THOMAS L A， et al. Identifying prominent life events on Twitter［C］// Proceedings of the 8th International Conference on Knowledge Capture. New York： ACM， 2015： No.4. 10.1145/2815833.2815845
3	CHOUDHURY S， ALANI H. Personal life event detection from social media［C］// Late‑breaking Results， Doctoral Consortium and Workshop Proceedings of the 25th ACM Hypertext and Social Media Conference. Aachen： CEUR‑WS.org， 2014： No.SP2014_04.
4	DHILLON P S， FOSTER D， UNGAR L. Multi‑view learning of word embeddings via CCA［C］// Proceedings of the 24th International Conference on Neural Information Processing Systems. Red Hook， NY： Curran Associates Inc.， 2011： 199-207.
5	KHODABAKHSH M， KAHANI M， BAGHERI E， et al. Detecting life events from twitter based on temporal semantic features［J］. Knowledge‑Based Systems， 2018， 148： 1-16. 10.1016/j.knosys.2018.02.021
6	LI J W， RITTER A， CARDIE C， et al. Major life event extraction from twitter based on congratulations/condolences speech acts［C］// Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing. Stroudsburg， PA： Association for Computational Linguistics， 2014： 1997-2007. 10.3115/v1/d14-1214
7	BLEI D M， NG A Y， JORDAN M I. Latent Dirichlet allocation［J］. Journal of Machine Learning Research， 2003， 3： 993-1022.
8	CHOUDHURY S， ALANI H. Detecting presence of personal events in Twitter streams［C］// Proceedings of the 2014 International Conference on Social Informatics， LNCS 8852. Cham： Springer， 2015： 157-166.
9	LI J W， CARDIE C. Timeline generation： tracking individuals on twitter［C］// Proceedings of the 23rd International Conference on World Wide Web. New York： ACM， 2014： 643-652. 10.1145/2566486.2567969
10	HOCHREITER S， SCHMIDHUBER J. Long short‑term memory［J］. Neural Computation， 1997， 9（8）： 1735-1780. 10.1162/neco.1997.9.8.1735
11	SCHUSTER M， PALIWAL K K. Bidirectional recurrent neural networks［J］. IEEE Transactions on Signal Processing， 1997， 45（11）： 2673-2681. 10.1109/78.650093
12	ZHOU C T， SUN C L， LIU Z Y， et al. A C‑LSTM neural network for text classification［EB/OL］. （2015-11-30）［2021-09-10］..
13	YEN A‑Z， HUANG H‑H， CHEN H‑H. Detecting personal life events from Twitter by multi‑task LSTM［C］// Proceedings of the 2018 Web Conference Companion. Republic and Canton of Geneva， Switzerland： International World Wide Web Conferences Steering Committee， 2018： 21-22. 10.1145/3184558.3186909
14	YEN A Z， HUANG H H， CHEN H H. Personal knowledge base construction from text‑based lifelogs［C］// Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York： ACM， 2019： 185-194. 10.1145/3331184.3331209
15	YEN A Z， HUANG H H， CHEN H H. Multimodal joint learning for personal knowledge base construction from Twitter‑based lifelogs［J］. Information Processing and Management， 2020， 57（6）： No.102148. 10.1016/j.ipm.2019.102148
16	DEVLIN J， CHANG M W， LEE K， et al. BERT： pre‑training of deep bidirectional transformers for language understanding［C］// Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics： Human Language Technologies， Volume 1 （Long and Short Papers）. Stroudsburg， PA： ACL， 2019： 4171-4186. 10.18653/v1/n18-2
17	ZHOU P， SHI W， TIAN J， et al. Attention‑based bidirectional long short‑term memory networks for relation classification［C］// Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics （Volume 2： Short Papers）. Stroudsburg， PA： ACL， 2016： 207-212. 10.18653/v1/p16-2034
18	LAFFERTY J D， McCALLUM A， PEREIRA F C N. Conditional random fields： probabilistic models for segmenting and labeling sequence data［C］// Proceedings of the 18th International Conference on Machine Learning. San Francisco： Morgan Kaufmann Publishers Inc.， 2001： 282-289.

[1]	Ping LUO, Ling DING, Xue YANG, Yang XIANG. Chinese event detection based on data augmentation and weakly supervised adversarial training [J]. Journal of Computer Applications, 2022, 42(10): 2990-2995.
[2]	LI Yanzhi, FAN Yong, GAO Lin. Anomaly detection of oil drilling water flow based on shape flow [J]. Journal of Computer Applications, 2021, 41(6): 1842-1848.
[3]	MENG Xiangrui, YANG Wenzhong, WANG Ting. Survey of sentiment analysis based on image and text fusion [J]. Journal of Computer Applications, 2021, 41(2): 307-317.
[4]	Cui WANG, Yafei ZHANG, Junjun GUO, Shengxiang GAO, Zhengtao YU. Event detection without trigger words incorporating syntactic information [J]. Journal of Computer Applications, 2021, 41(12): 3534-3539.
[5]	GUO Kexin, ZHANG Yuxiang. Visual-textual sentiment analysis method based on multi-level spatial attention [J]. Journal of Computer Applications, 2021, 41(10): 2835-2841.
[6]	LI Shanshan, YANG Wenzhong, WANG Ting, WANG Lihua. Survey of sub-topic detection technology based on internet social media [J]. Journal of Computer Applications, 2020, 40(6): 1565-1573.
[7]	CAI Guoyong, HE Xinhao, CHU Yangyang. Visual sentiment analysis by combining global and local regions of image [J]. Journal of Computer Applications, 2019, 39(8): 2181-2185.
[8]	MAO Yingchi, QI Hai, JIE Qing, WANG Longbao. M-TAEDA: temporal abnormal event detection algorithm for multivariate time-series data of water quality [J]. Journal of Computer Applications, 2017, 37(1): 138-144.
[9]	PAN Lei, ZHOU Huan, WANG Minghui. Real-time detection method of abnormal event in crowds [J]. Journal of Computer Applications, 2016, 36(6): 1719-1723.
[10]	LU Zhigang, SUN Yadan. Multidimensional collaborative intelligence recommendation based on social media context [J]. Journal of Computer Applications, 2016, 36(3): 740-745.
[11]	CAI Guoyong, XIA Binbin. Multimedia sentiment analysis based on convolutional neural network [J]. Journal of Computer Applications, 2016, 36(2): 428-431.
[12]	MAO Yingchi, JIE Qing, CHEN Hao. Online abnormal event detection with spatio-temporal relationship in river networks [J]. Journal of Computer Applications, 2015, 35(11): 3106-3111.
[13]	Xiao-fei XUE Yong-kui ZHANG Xiao-dong REN. Method research of new event detection based on news element [J]. Journal of Computer Applications, 2008, 28(11): 2975-2977.
[14]	SUN Zhao-lin,YANG Hong-wen,HU Wei-dong. Methods based on the Bayesian networks for situation assessment [J]. Journal of Computer Applications, 2005, 25(04): 745-747.

Personal event detection method based on text mining in social media

基于社交媒体文本挖掘的个人事件检测方法

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 13

References 18

Related Articles 14

Recommended Articles

Metrics