摘要
随着基于地理位置服务的不断发展,在线地图应用(WMS)成为了人们生活中不可缺少的一部分.在PC端WMS大数据的基础上,以新颖的视角对用户的搜索行为和不同城市的搜索差异进行了测量、分析和理解.首先从宏观和微观两个角度对用户搜索时间进行分析,指出WMS数据不同于其他地理信息数据,具备搜索行为前瞻的特性;随后,验证了每个城市高频查询兴趣点的查询频次符合Zipf分布,并解释了分布参数所蕴含的物理意义;更进一步的,用简单直观的方法定量研究了城市之间的流动性和城市的人流模式.为随后的基于WMS数据的进一步挖掘和研究提供了测量基础.
With the rapid growth of location-based services （LBS） ,web map service （WMS） is becoming indispensable in our daily life. From a new perspective, this paper measures and analyzes the user behaviors and regional differences in WMS, based on a big log dataset from the PC clients of a large-scale WMS provider. We give analysis on users＇ searching times from both macro and micro per- spective, and point out that WMS data has a feature of searching behavior prediction, which is absent in other location-based datasets. Then, we observe and verify that the searching frequencies of point of interests in a city conform to Zipf distribution, and explain the underlying physical meanings of the corresponding parameters. In addition, we present a simple and intuitive approach to quantitatively study the inter-city fluidity and intra-city mobility patterns. And our work can serve as a measurement basis for future work in the area of WMS data mining.
出处
《小型微型计算机系统》
CSCD
北大核心
2015年第1期33-38,共6页
Journal of Chinese Computer Systems
基金
自然科学基金项目(61272459
61170245)资助
国家"八六三"计划(2013AA013501)资助
工信部重大专项(2013ZX03002003-004)资助
中央高校基本科研业务费
陕西省工业攻关项目(2013K06-38)资助
关键词
地理兴趣点
前瞻特性
ZIPF分布
城市流动性
web map service
point of interest
searching behavior prediction
Zipf distribution
inter-city fluidity
