论文部分内容阅读
针对电子中介网站中多属性搜索问题,提出一种基于用户对属性满意度的搜索推荐模型。首先,给出电子中介网站中用户使用多属性搜索的问题描述;应用满意度计算规则[2],考虑用户满意度最大为目标构建多目标搜索模型。以二手书电子中介为例,给出具有良好用户体验的推荐搜索列表。在此基础上,应用朴素贝叶斯算法[4,5],对网站高级用户的购买历史行为进行分析,根据用户对属性的偏好,给出特别推荐作为参考。
Aiming at the problem of multi-attribute search in electronic intermediary websites, a search recommendation model based on user’s satisfaction with attributes is proposed. First, the problem description of users using multi-attribute search in electronic intermediary websites is given. The algorithm of satisfaction calculation [2] is applied to construct multi-objective search model with the goal of maximizing user satisfaction. The second-hand book electronic intermediary, for example, gives a good user experience recommended search list. On this basis, the naive Bayes algorithm [4, 5] is applied to analyze the purchase history behavior of advanced users. Based on the user’s preference of attributes, special recommendations are given as reference.