摘要
针对如何有效实现个性化推荐服务的问题,在农业信息推荐系统的设计过程中,采用基于内容过滤的推荐技术,提出一种新的用户综合兴趣模型。模型通过将用户背景、阅读与操作行为等因素进行综合加权,计算用户与文档的相似度,并以此向用户推荐文档。测试结果表明,提高用户阅读与操作特征在模型中所占的权重,可以有效提高推荐精度。
To realize the personalized recommendatory service effectively,the content-based filtering recommendatory technology is adopted in the design process of the agriculture information recommendation system.A new user comprehensive interest model is presented.The model can calculate the similarity of user and the document by the synthetically weights of background,reading and operation behavior factors of users,and recommend document to users based on this.By the practical tests,the results show that increasing the weights of user’s reading and operating behavior in model can effectively improve the recommended precision.
出处
《计算机工程》
CAS
CSCD
2012年第11期38-41,共4页
Computer Engineering
基金
国家自然科学基金资助项目(60963014
60663307)
江西省自然科学基金资助项目(2007GZS0186)
关键词
农业信息
用户模型
推荐技术
相似度
内容过滤
权重因子
agricultural information
user model
recommendation technology
similarity
content filtering
weight factor
