The Study Of Spatio-temporal Similarity Of Vehicle Trajectory Data Based On The Clustering Algorithm

With development and update of new communication technology and data acquisition technology, Huge amounts of data constantly emerge. With the index increase and expansion of data, data mining and analysis become more and more important. What’s more, It’s great significance to guide industry, agriculture, commerce, medical treatment, human behavior, social production and life. In this paper for example of the Beijing taxi GPS trajectory of space and time data, by data mining, and one of the methods of data mining called clustering as the analysis tool, relevant information was extracted and dug. It is different from the traditional static analysis of data mining, which is combined with dynamic clustering and data mining of time, effectively get meaningful and semantic similarity model, then, introduces the applying of semantic model used in the production and living. Firstly, It’s started with the concept and theory of data mining and the related study of clustering algorithm, to master the meaning and processing method of data mining and to insight into relevant knowledge of algorithm steps, which provides the analysis of technical support for the realistic application of trajectory data analysis and mining of time and space. Afterwards, according to the process of data mining, vehicle GPS trajectory data of the typical trajectory data of space and time were dealt with by preprocessing noise reduction, time space parallel clustering, extracting semantic information application in the actual problem. This topic is about the exploration of future data analysis and data mining which are multidimensional, as well as relevant knowledge of overview of data mining and clustering analysis theory.