摘要
社交媒体中的文本内容可对交通量数据进行补充,为此提出一个交通事件可视分析方法。建立交通事件文本处理模型,提取事件的描述信息;基于图嵌入算法学习道路节点属性的向量表示,建立道路相似性模型;结合核密度模型建立交通事件发生概率预测模型;设计了一个交互式可视分析界面对于交通事件进行可视分析与探索。通过交通信息抽取、道路相似性度量以及交通事件交互预测等案例分析,验证了所提方法的有效性,可以辅助交通部门管理决策。
Traffic text data in social media can supplement traffic flow information,for which a visual analysis method for traffic events is proposed.A text processing model is designed to process the social media data and extract the description information of traffic event.The vector representation of road node attributes is learned based on graph embedding algorithm,and a road similarity model is estbalished.A prediction model of the traffic event is built based on the road similarity and the kernel density model.An interactive visual analysis system is designed to carry out visual analysis.Though traffic information extraction,road similarity measurement and traffic event interaction prediction,the effectiveness of the proposed method is verified and can assist traffic department management decisions.
作者
吴向平
平力俊
徐懂事
Wu Xiangping;Ping Lijun;Xu Dongshi(College of Information Engineering,China Jiliang University,Hangzhou 310018,China;Key Laboratory of Electromagnetic Wave Information Technology and Metrology of Zhejiang Province,China Jiliang University,Hangzhou 310018,China)
出处
《系统仿真学报》
CAS
CSCD
北大核心
2022年第5期1140-1151,共12页
Journal of System Simulation
基金
浙江省基础公益研究计划(LGF20F020012)。
关键词
数据可视化
信息抽取
相似性分析
交通事件预测
可视分析
data visualization
information extraction
road similarity
traffic prediction
visual analytics
