传播力约束下的微地图用户关联规则挖掘算法  

Mining algorithm of We-map users association rule under the constraint of communication capacity

作　　者：张剑 闫浩文 王卓 白娅兰 ZHANG Jian;YAN Haowen;WANG Zhuo;BAI Yalan(Faculty of Geomatics,Lanzhou Jiaotong University,Lanzhou 730070,China;Nation-Local Joint Engineering Research Center of Technologies and Applications for National Geographic State Monitoring,Lanzhou 730070,China;Gansu Provincial Engineering Laboratory for National Geographic State Monitoring,Lanzhou 730070,China)

机构地区：[1]兰州交通大学测绘与地理信息学院,兰州730070 [2]地理国情监测技术应用国家地方联合工程研究中心,兰州730070 [3]甘肃省地理国情监测工程实验室,兰州730070

出　　处：《测绘科学》2022年第1期227-235,共9页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(41671447);兰州交通大学优秀平台支持项目(201806)。

摘　　要：针对微地图在自媒体时代的传播特性,以传播力指数作为关联规则挖掘的衡量标准,设计出了一种符合微地图用户行为的关联规则挖掘算法。首先,分析微地图传播力的影响因子,构建微地图传播力评估体系,并利用加权平均的方法对微地图传播力指数进行量化。然后,结合经典的关联规则FP-growth算法,提出了一种传播力约束下的行为关联规则算法。其中,将用户信息和传播力指数进行排序、映射后,构建"传播力指数模式树",在保留数据中关联信息的同时可直接进行关联规则挖掘。最后,将"传播力指数"与"支持度-置信度"相结合,设置阈值最小传播力指数,挖掘传播力强的用户行为关联组合。实验表明,该算法可以筛选出更能表现大众传媒习惯的关联规则,对强传播力组合的用户行为习惯有较好的挖掘效果。这使得微地图用户的行为更加清楚,微地图的推荐系统更加合理,为大众传媒下的关联规则算法提供了一种思路。Aiming at the communication characteristics of We-map in the self-media era, this paper utilizes the communication capacity index as the measurement standard for association rule mining to design an association rule mining algorithm that conforms to the user’s behavior of We-map. First, this paper analyzes the influencing factors of the We-map communication capacity, constructs a We-map communication capacity evaluation system, and uses the weighted average method to quantify the We-map communication capacity index. Then, combining with the classic association rule FP-growth algorithm, a behavioral association rule algorithm under the constraint of communication capacity is proposed. Among them, the user information and the communication capacity index are sorted and mapped to construct a “communication capacity index pattern tree”,which can directly perform association rule mining while retaining the associated information in the data. Finally, combining the “Communication Capacity Index” with the “Support Degree-Confidence Degree” to set a threshold of minimum communication capacity index and mine the combination of user behaviors with strong communication capacity. Experiments show that the algorithm can filter out association rules that can better express the habits of mass media, and has a better mining effect on user behavior and habits with strong communication capacity combinations. The algorithm makes the behavior of We-map users clearer and the recommendation system of We-map more reasonable, which provides a way of thinking for the association rule algorithm under the mass media.

关 键 词：微地图 传播力评估体系 传播力指数 传播力指数模式树 关联规则挖掘 

分 类 号：P[天文地球]