基于实体匹配的互联网餐馆POI数据整合研究

Research on Internet Restaurant POI Data Integration Based on Entity Matching

针对当前互联网地图中POI语义信息现势性差的问题,以南昌市餐馆POI数据为研究对象,采用基于属性与距离的实体匹配方法,对百度地图餐馆POI数据和美团网餐馆数据进行整合研究。利用596条百度餐馆数据和747条美团网餐馆数据进行实验,得到了127条语义信息完备且实时性强的餐馆POI数据。实验结果表明,本文方法可行,为互联网地图的动态更新提供了一种新的途径。整合后的POI数据可以为公众日常出行就餐、商业规划等提供强有力的数据支撑。

In view of the poor up-to-dateness of POI semantic information in the current internet map,taking the POI data of Nanchang restaurant as the research object,this paper uses the entity matching method based on attribute and distance to integrate Baidu map restaurant POI data and Meituan restaurant data.Using 596 Baidu map restaurant data and 747 Meituan restaurant data,127 restaurant POI data with complete semantic information and strong real-time are obtained.The results show that the method is feasible and provides a new way for the dynamic updating of internet map.The integrated POI data can provide strong data support for public daily travel,dining and business planning.