Web环境下地学数据共享用户行为模式分析

发布时间：2018-06-19 05:25

本文选题：网络数据挖掘 + 空间数据挖掘　；参考：《地球信息科学学报》2016年09期

【摘要】：了解科学数据共享用户行为特征对实现高效、精准的数据共享服务具有重要的参考意义。本文基于国家地球系统科学数据共享平台网站服务器日志及服务记录数据,利用空间数据挖掘及Web使用挖掘技术,探索地球系统科学数据共享用户行为模式。在数据预处理阶段,完成用户识别、会话识别、位置识别,并对数据进行空间建模、空间数据库建库。在数据挖掘阶段,分别对用户产生的网页浏览数、会话数、数据集浏览数为对象进行空间"热点"分析,识别用户行为的地域差异。针对用户数据浏览和下载行为,采用FP-growth算法对用户——数据之间进行关联规则挖掘,发现用户对数据关注和使用的高频规律。分析结果表明:(1)该共享平台用户地在国内各省市均有分布,用户最多的3个省(市)分别为北京市、山东省、江苏省,该分布与国内高校学生分布相关程度不高,但与"211工程"高校学生的空间分布相关度较高;(2)空间"热点"分析表明,北京、天津及河北北部无论在网页浏览、数据浏览还是会话量上都是"热点"区域,但识别的"冷点"区域有较大不同,尤其是数据访问"冷点"分布较广,如南方沿海省份、河南省、山东省、四川省等;(3)关联规则挖掘发现多个数据浏览高频项目集以及关联规则。数据下载高频项与数据浏览高频模式较好吻合,但下载行为未表现出明显关联规则。本文提供了一种结合Web使用挖掘和空间数据挖掘的用户行为模式挖掘方法,该方法也可用于其他类型网站的数据挖掘。
[Abstract]:Understanding the behavior characteristics of scientific data sharing users has important reference significance for the realization of efficient and accurate data sharing services. Based on the web server log and service record data of National Earth system Science data sharing platform, this paper explores the user behavior pattern of Earth system Science data sharing by using spatial data mining and Web usage mining technology. In the stage of data preprocessing, user identification, session recognition, location recognition are completed, and the spatial modeling of the data is carried out, and the spatial database is built. In the stage of data mining, the spatial "hot spot" analysis is carried out to identify the regional differences of users' behavior, including the number of page views, the number of sessions and the number of views of data sets generated by users. According to the behavior of user data browsing and downloading, FP-growth algorithm is used to mine association rules between users and data, and the high frequency rules of users' attention and use of data are found. The results show that the shared platform is distributed in all provinces and cities in China, and the three provinces (cities) with the most users are Beijing, Shandong and Jiangsu respectively. The degree of correlation between this distribution and the distribution of college students in China is not high. However, the spatial "hot spot" analysis of Beijing, Tianjin and northern Hebei shows that Beijing, Tianjin and northern Hebei are "hot spots" in terms of web browsing, data browsing and conversation volume. But the "cold spot" areas identified are quite different, especially the "cold spots" of data access, such as southern coastal provinces, Henan Province, Shandong Province, The mining of association rules finds many data browsing high frequency itemsets and association rules. The data download high frequency item is in good agreement with the data browsing high frequency mode, but the downloading behavior does not show obvious association rules. This paper presents a user behavior pattern mining method combining Web usage mining and spatial data mining, which can also be used in other types of Web sites.
【作者单位】：中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室;中国科学院大学;江苏省地理信息资源开发与利用协同创新中心;
【基金】：国家科技基础条件平台——地球系统科学数据共享平台(2005DKA32300) 科技基础性工作重点项目(2011FY110400) 中国工程院国际工程科技知识中心项目
【分类号】：TP311.13

【相似文献】