基于本体相似度和时间衰减的动态个性化推荐算法

发布时间：2018-05-19 22:21

本文选题：个性化 + 推荐　；参考：《图书情报工作》2017年S1期

【摘要】：为解决当前个性化推荐不能同时考虑用户兴趣点受时间和频率两个维度共同影响问题,提出一种基于本体相似度和时间衰减的动态个性化推荐算法。首先确定时间衰减函数来计算用户兴趣点的时间衰减规律,然后考虑不同兴趣点访问频率对兴趣点关注程度的影响,再次通过本体相似的综合算法解算推荐资源与关注目标资源的相似度,最后通过兴趣点更新权重与相似度的加成效果确定推荐资源的排序。实验结果表明该算法在整体上具有较好的综合性能,推荐资源的查准率和查全率均较高,算法的普适性和效果较好。
[Abstract]:A dynamic personalized recommendation algorithm based on ontology similarity and time attenuation is proposed to solve the problem that users' interest points can not be affected by both time and frequency dimensions. First, the time attenuation function is determined to calculate the time decay law of the user's point of interest, and then the influence of different access frequency of the point of interest on the interest degree is considered. Thirdly, the similarity between the recommended resources and the target resources is calculated by the ontology similarity synthesis algorithm. Finally, the order of the recommended resources is determined by the addition effect of updating the weights and similarity of the points of interest. The experimental results show that the algorithm has better comprehensive performance on the whole, the recall rate and recall rate of recommended resources are both higher, and the universality and effect of the algorithm are better.
【作者单位】：云南财经大学科研院所党委;
【基金】：国家自然科学基金“综合交通体系下边境口岸物流均衡智能方法研究”(项目编号:61163035)研究成果之一
【分类号】：G252;TP391.3

【相似文献】