A study on knowledge acquisition from a large-scale traffic data set by using structured data mining techniques

本研究的目的是将结构数据挖掘技术应用于信息通信系统记录的大规模交通数据中，以获取交通规划所需的信息，并建立自动提取特征运动模式的分析技术。在本研究中，我们尝试运用各种数据挖掘技术对描述城市中人的运动的统计数据和城市中新的交通方式社区自行车（城市地区的自行车共享系统）的日志数据的人的出行数据进行挖掘。实现了对大规模数据集中某一特征运动模式估计的修正知识提取、高精度交通选择模型的构建、日常交通和非日常交通交通模式变化等整体交通现象的研究。

The aims of this study are to apply the structure data mining techniques for large-scale traffic data recorded by an information and communication system to obtain information to contribute to a transportation planning and to build analysis technique to extract a characteristic movement pattern automatically. In this study, we tried to apply various data mining techniques for the person trip data which were the statistics data described the movement of the person in the city and log data of community cycle (bicycle sharing system in an urban area) which was the new transportation mode in the city. As a result, the extraction of the knowledge to modify the estimate of a characteristic movement pattern to be in large-scale data set, the construction of a highly precise transportation choice model, the whole traffic phenomenon such as the change in the traffic pattern at the daily traffic and the non-daily traffic was enabled.