摘要
为提高互联网信息资源用户的服务满意度,对互联网信息资源用户获取进行优化推送。进行用户获取推送时,应依据用户浏览行为将其兴趣度的作用划分为正、逆两种方向,分析互联网用户兴趣度属性之间的依赖关系对信息推送问题建模,但是传统方法通过分析移动互联网通知消息和用户历史行为,筛选过滤对所推送互联网信息感兴趣的用户群完成推送,但是不能对用户兴趣度进行有效划分,无法准确分析用户兴趣度属性之间的依赖关系,导致推送结果不可靠,推送形式和推送信息类型较为单一的问题,提出一种基于MQTT Proxy的互联网信息资源用户获取优化推送方法。首先获取MQTT Proxy下移动互联网用户的历史浏览行为,计算出用户的不同历史浏览行为特征权重,融合于熵权理论,依据用户浏览行为将其兴趣度的作用划分为正、逆两种方向,计算出移动互联网用户各种浏览行为的信息熵值,以此为依据组建基于MQTT Proxy的互联网用户兴趣度模型。其次结合朴素贝叶斯理论,通过分析MQTT Proxy的互联网用户兴趣度属性之间的依赖关系对信息推送问题建模,从中度量互联网共邻用户兴趣偏好之间关系对推送用户对之间的贡献和影响,计算其近似度并进行排序,由此完成互联网信息资源用户获取优化推送。仿真结果表明,所提方法能够有效提升互联网信息资源用户获取推送准确性,获得更好的信息推送效果。
An optimization push method for user acquisition of intemet information resource based on the MQTr Proxy is proposed. Firstly, the historical browsing behaviors of mobile internet user under MQTT Proxy are acquired, and different feature weight of user historical browsing behaviors are worked out. Then, the effect of user interest de- gree is divided into the positive and reverse directions according to user browsing behavior integrated with the entropy weight theory, and the information entropy of various user browsing behaviors are worked out. The interest degree model of internet user based on the MQTI＂ Proxy is built. Moreover, the model for the information push is built through analyzing the dependence relationship among user interest attributes. The effect and contribution of relation- ship between the interest and preference of common adjacency user is measured. Finally, its degree of approximation is calculated and ranked. Thus, the optimization push for user acquisition of interact information resource is comple- ted. Simulation results show that the proposed method can improve the push accuracy and obtain good information push effect.
出处
《计算机仿真》
北大核心
2017年第7期273-276,280,共5页
Computer Simulation
