情境感知和兴趣适应的农业信息推荐模型  

Context-aware and interest adaptation agricultural information recommendation model

作　　者：张新香[1] 

机构地区：[1]中南财经政法大学信息与安全工程学院,武汉430073

出　　处：《计算机应用研究》2016年第5期1315-1318,共4页Application Research of Computers

基　　金：国家教育部人文社科青年基金资助项目(12YJCZH286);国家自然科学基金资助项目(71373286);2011年度中央高校基本科研业务费青年教师资助项目(31541111306)

摘　　要：个性化推荐是解决农户需求和农业信息服务契合度的关键技术。传统基于行为的推荐算法能挖掘农户兴趣但不能反映其兴趣的迁移,时间权重的引入较好地实现了自适应机制。地域、时间以及农业生产门类等情境信息直接影响农户的兴趣偏好,已有算法却融入度不足。综合情境因素、浏览行为和兴趣改变等多维因素,拓展向量空间模型,设计情景感知和兴趣适应的农业信息推荐算法。实验结果显示,改进后的算法能显著提升推荐的准确率和有效性。Personal recommendation was a key technology to match farmers ＇ requirements and agricultural information services. Traditional recommendation algorithms based on browsing behaviors could mine users＇ interests,but could not reflect the change of users＇ interests. The algorithms incorporated with time-based data weight could realize self-adaptive mechanism.Contextual information such as region,time,and agricultural production categories were relevant with farmers＇ preferences.However,the existing algorithms seldom involved the contextual information. Synthesizing several factors including context,browsing behaviors and interest change,based on the vector space model（ VSM）,this paper proposed a context-aware and interest adaptation agriculture information recommendation algorithm. Experimental results show that the proposed algorithm can efficiently improve the recommendation precision and quality.

关 键 词：农业信息 个性化推荐 情境感知 兴趣适应 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]