Please wait a minute...
Data Analysis and Knowledge Discovery  2020, Vol. 4 Issue (12): 136-147    DOI: 10.11925/infotech.2096-3467.2020.0596
Current Issue | Archive | Adv Search |
Identifying Traffic Events from Weibo with Knowledge Graph and Target Detection
Sun Xinrui1,Meng Yu2,Wang Wenle1()
1School of Software, Jiangxi Normal University, Nanchang 330022, China
2School of Mathematics Science, Qufu Normal University, Qufu 273100, China
Download: PDF (3022 KB)   HTML ( 11
Export: BibTeX | EndNote (RIS)      
Abstract  

[Objective] This paper identifies traffic events from Weibo (microblog) posts with the help of knowledge graph and target detection techniques, aiming to address traffic management issues. [Methods] First, we constructed traffic knowledge graph and event evolution graph based on open data. Then, we identified traffic events from microblog texts. Third, we retrieved microblog images with target detection to further improve the recognition accuracy of three types of events. [Results] We examined our method with microblog data on Qingdao’s traffics in 2018. The precision of traffic event detection based on texts and images were 94.55% and 95.53%. [Limitations] More research is needed to reduce the manual construction of traffic knowledge graph, and improve the target detection algorithm. [Conclusions] The proposed method could help urban traffic management departments detect road incidents or traffic problems, and then facilitate their decision-makings.

Key wordsIntelligent Traffic System (ITS)      Traffic Knowledge Graph      Traffic Event Identification      Traffic Target Detection     
Received: 23 June 2020      Published: 25 December 2020
ZTFLH:  G350  
Corresponding Authors: Wang Wenle     E-mail: wenlewang@jxnu.edu.cn

Cite this article:

Sun Xinrui,Meng Yu,Wang Wenle. Identifying Traffic Events from Weibo with Knowledge Graph and Target Detection. Data Analysis and Knowledge Discovery, 2020, 4(12): 136-147.

URL:

http://manu44.magtech.com.cn/Jwk_infotech_wk3/EN/10.11925/infotech.2096-3467.2020.0596     OR     http://manu44.magtech.com.cn/Jwk_infotech_wk3/EN/Y2020/V4/I12/136

Technical Roadmap
事件类别 基础要素 补充信息 计算指标
交通拥堵 位置、时间、拥堵特征表述(词、短句) 程度、原因、建议 完备度、拥堵指数、可信度
交通事故 位置、时间、事故特征表述 涉事车辆/人物、程度、后果 完备度、严重指数、可信度
交通反馈 位置、时间、反馈对象、对象状态/行为 后果 完备度、紧急指数、可信度
Traffic Event Extraction Template
Schematic Diagram of Basic Construction Process of Traffic Knowledge Graph
顶层节点 包含实体 实体属性 数据来源
司机、行人、交警、路段养护人员、商贩等 性别、年龄段等 自定义
按用途分类、按行驶道路条件分类 通用属性,包括大小、颜色、别称等,进一步细化到车的品牌、型号等 《中国汽车分类标准(GB9417-89)》、
大型汽车垂直网站
城市市区道路 路名、类型、交叉路、方向、车道数、所属区域 城市公开数据、交通管理部门提供数据
信息点(Point of Information, POI) 名称、所属道路、所属区域 中国POI数据网采集
环境 天气 名称、类别 天气种类自定义
道路交通安全设施 名称、类别、作用、正常状态、异常状态 交通百科
Four Elements of Traffic Knowledge Graph
Basic Event Chain in Traffic Event Graph
Event Extraction Process of Traffic Microblog
事件类型 表述形式 示例 标注
交通拥堵 直接表述-关键词 堵车、堵了、拥堵 EV-CON-D-S
直接表述-状态短语 堵成、堵的、堵到 EV-CON-D-S
间接表述-指代词 交通瘫痪、交通压力 EV-CON-I-S
交通事故 直接表述-关键词(名) 交通事故、车祸 EV-ACC-KWN
直接表述-关键词(动) 追尾、碰撞、刮蹭、撞 EV-ACC-KWV
间接表述-示意词 吻了、亲密接触 EV-ACC-I-NS
交通反馈 直接表述-基础事件词 乱停车、违停、闯红灯 EV-FD-KW-BH
Traffic Incident Triggers’ Identification and Marking
地点要素 表述形式 示例词 标注
路名 全称 山东路、香港路 LO-RD
简称 海尔桥、杭鞍高架、杭鞍 LO-RD-S
别称 东西快速路->杭鞍高架路 LO-RD-B
定位信息 附属定位词 交叉口、入口、路口 LO-CW
车道 直行车道、左转车道 LO-RD-LA
POI名称 鞍山路小学、市图书馆 LO-POI
距离描述词 一公里、100米 LO-DIS
方向 双向、北向南、东往西 LO-RD-DI
Location Information’s Identification and Marking
表述形式 示例词 标注
时间点 9点10分、9:30、八点半 TIME-CP
时间段-一天不同时段 上午、早、晚上 PERIOD-DAY
时间段-日期 7月5日、今天、昨日 PERIOD-DATE
时间点-指代表述 这个点、此时、这个时候 TIME-RW
时间段-持续时间 半个小时、40多分钟、100秒 PERIOD-HMS
时间段-模糊表述 早高峰、高峰时段 TIME-PE-NC
Time Information’s Identification and Marking
抽取要素 表述形式 示例词 标注
车信息 车型 公交车、轿车、SUV VE-TP
大小 大车、小车 VE-SIZE
品牌、型号 奔驰车、斯柯达、奥迪A4 VE-MB
颜色 红车、白车 VE-CO
车牌 鲁B12345 VE-PL
数量词 量词+车 两车、一辆 VE-NUM
Vehicle Elements’ Identification and Marking
环境元素 表述形式 示例词 标注与抽取方法
天气 形容词+天气 好天气、下雨天 EN-WE-DW
具体天气词 大雨、大雾、大雪、雾霾 EN-WE-TP
动词+天气词 下雨、下雪 EN-WE
交通安全设施 信号灯 信号灯、红绿灯、红灯 EN-TL
交通标志 禁停标志、直行标志 EN-TS
交通标线 斑马线、黄色网格线 EN-TLM
道路护栏 护栏、栏杆、隔离 EN-GB
路面状态 直接表述 路面破损 EN-RDSF
间接表达-状态 坑坑洼洼、积水深 EN-RDSF-DW
其他 直接表述 电线杆、天桥、花坛 EN-OTE
Environment Elements’ Identification and Marking
人物元素 表述形式 示例词 标注
司机 车+司机 公交车司机、宝马司机 PE-DR-VE
描述信息+司机 男司机、女司机、无德司机 PE-DR-DE
行人 群指-无明确个人 行人、路人 PE-CP
个指-有描述信息 老人、小孩、学生 PE-CH
交警 原称 交警、警察 PE-PO
别称 交警蜀黍、焦警 PE-PO-B
其他 道路服务人员 施工人员、清洁工 PE-RS
商贩 商贩、摊贩 PE-SE
行乞人员 乞讨者 PE-BE
Character Elements’ Identification and Marking
Target Detection Data
关键词组合方式 具体内容
城市名+路名 青岛+1180条青岛市区道路名称
城市名+交通种子词 青岛+8个基础道路交通事件特征词(堵、事故、车祸、追尾、撞、信号灯、红绿灯、乱停车)
@+城市交通官方账号 20个青岛交通官方微博账号(青岛交通广播FM897、青岛交警、青岛邱磊、市北交警、李沧交警、青岛发布、青岛公安、市南交警、青岛随手拍、半岛都市报等)
Key Words Combination of Data Collection
事件类别 完备事件数 独立事件数
交通拥堵 2 991 1 826
交通事故 1 046 680
交通反馈 2 463 1 450
总计 6 500 3 956
Traffic Event Results Based on Traffic Knowledge Graph Extraction
事件类别 准确率/% 召回率/% F1值/%
交通拥堵 96.46 94.58 95.51
交通事故 94.31 93.27 93.79
交通反馈 92.89 90.51 91.68
总计 94.55 92.79 93.66
Test Results of Traffic Incident Detection
检测目标 准确率/% 召回率/% F1值/%
轿车(Car) 92.61 88.67 90.60
公交车(Bus) 93.42 85.94 89.52
卡车(Truck) 92.17 87.69 89.87
货车(Van) 90.33 89.91 90.12
摩托车(Motorbike) 95.29 90.64 92.91
行人(Person) 96.47 87.58 91.81
信号灯(Traffic Light) 91.58 90.13 90.85
Test Results of Traffic Target Element Detection
Image Detection Examples
事件类型 准确率/% 召回率/% F1值/%
交通拥堵 96.64 95.19 95.91
交通事故 94.88 92.83 93.84
交通反馈 95.07 91.33 93.16
总计 95.53 93.12 94.30
Experimental Result
事件类型 应识别数 正确
识别数
准确率/
%
召回率/
%
F1值/
%
交通拥堵 50 47 100.00 94.00 96.91
Experimental Results of Incomplete Traffic Congestion Event Picture and Event Detection
[1] Rosi A, Mamei M, Zambonelli F, et al. Social Sensors and Pervasive Services: Approaches and Perspectives [C]//Proceedings of 2011 IEEE International Conference on Pervasive Computing and Communications Workshops. IEEE, 2011: 525-530.
[2] Misra A, Gooze A, Watkins K , et al. Crowdsourcing and Its Application to Transportation Data Collection and Management[J]. Transportation Research Record: Journal of the Transportation Research Board, 2014,2414(1):1-8.
[3] Wang X, Zheng X H, Zhang Q P , et al. Crowdsourcing in ITS: The State of the Work and the Networking[J]. IEEE Transactions on Intelligent Transportation Systems, 2016,17(6):1596-1605.
[4] Wang F Y . The Emergence of Intelligent Enterprises: From CPS to CPSS[J]. IEEE Intelligent Systems, 2010,25(4):85-88.
[5] Wang F Y . Real-Time Social Transportation with Online Social Signals[J]. IEEE Transactions on Intelligent Transportation Systems, 2014,15(3):909-914.
[6] 王飞跃 . 社会信号处理与分析的基本框架:从社会传感网络到计算辩证解析方法[J]. 中国科学:信息科学, 2013,43(12):1598-1611.
[6] ( Wang Feiyue . A Framework for Social Signal Processing and Analysis: From Social Sensing Networks to Computational Dialectical Analytics[J]. Chinese Science: Information Science, 2013,43(12):1598-1611.)
[7] Wang F Y . Crowdsourcing for Field Transportation Studies and Services[J]. IEEE Transactions on Intelligent Transportation Systems, 2015,16(1):1-8.
[8] Endarnoto S K, Pradipta S, Nugroho A S, et al. Traffic Condition Information Extraction & Visualization from Social Media Twitter for Android Mobile Application [C]//Proceedings of the 2011 International Conference on Electrical Engineering and Informatics. IEEE, 2011: 1-4.
[9] Gutiérrez C, Figuerias P, Oliveira P, et al. Twitter Mining for Traffic Events Detection [C]//Proceedings of 2015 Science and Information Conference. IEEE, 2015: 371-378.
[10] Wang S Z, He L F, Stenneth L, et al. Citywide Traffic Congestion Estimation with Social Media [C]//Proceedings of the 23rd SIGSPATIAL International Conference on Advances in Geographic Information Systems. 2015.
[11] Zheng X H, Chen W, Wang P , et al. Big Data for Social Transportation[J]. IEEE Transactions on Intelligent Transportation Systems, 2016,17(3):620-630.
doi: 10.1109/TITS.2015.2480157
[12] Mai E, Hranac R. Twitter Interactions as a Data Source for Transportation Incidents [C]//Proceedings of Transportation Research Board 92nd Annual Meeting. 2013.
[13] D’Andrea E, Ducange P, Lazzerini B , et al. Real-Time Detection of Traffic from Twitter Stream Analysis[J]. IEEE Transactions on Intelligent Transportation Systems, 2015,16(4):2269-2283.
[14] Chen Y Y, Lv Y S, Wang X, et al. A Convolutional Neural Network for Traffic Information Sensing from Social Media Text [C]//Proceedings of the 20th International Conference on Intelligent Transportation Systems. IEEE, 2017.
[15] Zhang Z H, He Q, Gao J , et al. A Deep Learning Approach for Detecting Traffic Accidents from Social Media Data[J]. Transportation Research Part C: Emerging Technologies, 2018,86:580-596.
[16] 熊佳茜 . 基于CRF的中文微博交通信息事件抽取[D]. 上海: 上海交通大学, 2014.
[16] ( Xiong Jiaxi . Civil Transportation Event Extraction from Chinese Microblogs Based on CRF[D]. Shanghai: Shanghai Jiao Tong University, 2014.)
[17] 仇培元, 张恒才, 余丽 , 等. 微博客蕴含交通事件信息抽取的自动标注方法[J]. 中文信息学报, 2017,31(2):107-116.
[17] ( Qiu Peiyuan, Zhang Hengcai, Yu Li , et al. Automatic Event Labeling for Traffic Information Extraction from Microblogs[J]. Journal of Chinese Information Processing, 2017,31(2):107-116.)
[18] 郑治豪, 吴文兵, 陈鑫 , 等. 社交媒体大数据的交通感知分析系统[J]. 自动化学报, 2018,44(4):656-666.
[18] ( Zheng Zhihao, Wu Wenbing, Chen Xin , et al. A Traffic Sensing and Analyzing System Using Social Media Data[J]. Acta Automatica Sinica, 2018,44(4):656-666.)
[19] Grant-Muller S M, Gal-Tzur A, Minkov E , et al. Enhancing Transport Data Collection Through Social Media Sources: Methods, Challenges and Opportunities for Textual Data[J]. IET Intelligent Transport Systems, 2015,9(4):407-417.
doi: 10.1049/iet-its.2013.0214
[20] Redmon J, Farhadi A . YOLOv3: An Incremental Improvement[OL]. arXiv Preprint, arXiv: 1804.02767.
[1] Wang Yi,Shen Zhe,Yao Yifan,Cheng Ying. Domain-Specific Event Graph Construction Methods:A Review[J]. 数据分析与知识发现, 2020, 4(10): 1-13.
[2] Du Jian. Measuring Uncertainty of Medical Knowledge: A Literature Review[J]. 数据分析与知识发现, 2020, 4(10): 14-27.
[3] Nie Lei,Fu Juan,Yi Chengqi,Yang Daoling. Measuring Enterprise’s Offline Resumption with Mobile Device Positioning Data[J]. 数据分析与知识发现, 2020, 4(7): 38-49.
[4] Yue Lixin,Liu Ziqiang,Hu Zhengyin. Evolution Analysis of Hot Topics with Trend-Prediction[J]. 数据分析与知识发现, 2020, 4(6): 22-34.
[5] Pan Youneng,Ni Xiuli. Recommending Online Medical Experts with Labeled-LDA Model[J]. 数据分析与知识发现, 2020, 4(4): 34-43.
[6] Liang Yanping,An Lu,Liu Jing. Topic Resonance of Micro-blogs on Similar Public Health Emergencies[J]. 数据分析与知识发现, 2020, 4(2/3): 122-133.
[7] Deng Jiangao,Zhang Xuan,Fu Zhu,Wei Qingming. Tracking Online Public Opinion Based on System Dynamics: Case Study of “Xiangshui Explosion Accident”[J]. 数据分析与知识发现, 2020, 4(2/3): 110-121.
[8] Zhe Hu,Xianjin Zha,Yalan Yan. Interactive Behaviors of Online Health Community Users in Emergency[J]. 数据分析与知识发现, 2019, 3(12): 10-20.
[9] Guanghui Ye,Jinqing Yang. Route Recommendation Based on Two-way Link Analysis of Urban Name Entities[J]. 数据分析与知识发现, 2019, 3(11): 79-88.
[10] Hongfei Ling,Shiyan Ou. Review of Automatic Labeling for Topic Models[J]. 数据分析与知识发现, 2019, 3(9): 16-26.
[11] Bowen Liu,Rujiang Bai,Yanting Zhou,Xiaoyue Wang. Identifying Frontier Topics from Funding and Paper——Case Study of Carbon Nanotube[J]. 数据分析与知识发现, 2019, 3(8): 114-122.
[12] Xiuxian Wen,Jian Xu. Research on Product Characteristics Extraction and Hedonic Price Based on User Comments[J]. 数据分析与知识发现, 2019, 3(7): 42-51.
[13] Shiqi Deng,Liang Hong. Constructing Domain Ontology for Intelligent Applications: Case Study of Anti Tele-Fraud[J]. 数据分析与知识发现, 2019, 3(7): 73-84.
[14] Feng Guoming,Zhang Xiaodong,Liu Suhui. Classifying Chinese Texts with CapsNet[J]. 数据分析与知识发现, 2018, 2(12): 68-76.
[15] Chen Yunwei,Zhang Ruihong. Comparing on Community Detection Algorithms for Information Mining[J]. 数据分析与知识发现, 2018, 2(10): 84-94.
  Copyright © 2016 Data Analysis and Knowledge Discovery   Tel/Fax:(010)82626611-6626,82624938   E-mail:jishu@mail.las.ac.cn