计算机与现代化

• 数据库与数据挖掘 • 上一篇    下一篇

一种新的融合情景的美食推荐算法

  

  1.  1.南京理工大学计算机科学与工程学院,江苏南京210094;2.上海理工大学环境与建筑学院,上海200093
  • 收稿日期:2015-01-07 出版日期:2015-07-23 发布日期:2015-07-28
  • 作者简介:赵雪美(1988-),女,江苏徐州人,南京理工大学计算机科学与工程学院硕士研究生,研究方向:推荐系统,可信计算; 郭林锋(1988-),男,江苏南通人,上海理工大学环境与建筑学院硕士研究生,研 究方向:土木工程; 卞雪雯(1992-),女,江苏泰州人,硕士研究生,研究方向:可信计算。

A Food Recommendation Algorithm with Context

  1. 1. School of Computer Science and Engineering, Nanjing University of Science & Technology, Nanjing 210094, China;

    2. School of Environment and Architecture, Shanghai University of Science and Technology, Shanghai 200093, China
  • Received:2015-01-07 Online:2015-07-23 Published:2015-07-28

摘要:

针对传统协同过滤推荐算法不适用于情景因素,严重影响用户行为的这类场景,提出一种融合情景的推荐算法,并将该算法应用于美食推荐。首先,运用由情景属性构造向量表示情景,将情景信
息作为一个重要因素添加到兴趣模型中,从而产生UIC兴趣模型。根据用户在不同情景下使用方式的不同,重新创建当前用户与各情景相对应的子用户,得到以情景作为标识的用户项目评分矩阵。
针对融合情景的兴趣模型易产生数据稀疏问题,设计利用改进的WSlopeOne算法对未知评分进行填充;并通过对相似度公式进行优化,进而更加准确地找到当前用户的近邻,为用户提供更加有效的推荐
服务。最后,通过实验验证该算法的有效性。

关键词: 兴趣模型, 推荐算法, 协同过滤, 情景信息, 相似度公式

Abstract:

As the traditional collaborative filtering recommendation algorithm didn’t consider the situation that context information affected the users’ behavior seriously, a
recommendation algorithm with context was put forward and the algorithm was applied to food recommendation. First of all, the context was added to the traditional useritem
model expressed as an attributes vector, resulting in a UIC interest model. Then SubUsers were created according to different context from one user, thus obtaining a
new useritem ratings matrix in a certain context. Because the data sparseness problem was easy to generate in this approach, WSlopeOne algorithm was designed to predict
unknown ratings. Based on optimized similarity formula, more effective recommendation service would be provided that can more accurately find the current users neighbor,
providing users with good service. Last, experiments were done to verify the contents this paper proposed and expectation for further research was brought forward.

Key words: interest model, recommendation algorithm, collaborative filtering, context information, similarity formula

中图分类号: