一种考虑用户评分标准不一致性的在线商品评价方法
发布时间：2018-12-06 17:07
【摘要】：用户评分标准不一致导致不同用户对同一商品评分不可比较.为在不同用户对同一商品评分不可比较时提高在线商品评价的客观性,提出了一种考虑用户评分标准不一致性的在线商品评价方法.首先基于用户之间的相似性对不完全评分数据进行填充;然后根据填充得到的用户-商品评分矩阵计算用户对在线商品的偏好关系;最后基于群决策理论中的Schulze社会选择函数,通过构建加权有向图并计算最强路径得到商品评价.评价结果反映群体的偏好,从而使得消费者可以参考该评价结果做出正确的购买决策.同时,方法基于在线商品两两比较的评价方式,提高了操纵商品评价结果的复杂性.通过理论分析和实验验证了该商品评价方法的合理性和有效性.
[Abstract]:The inconsistency of the user rating standard leads to the incomparable rating of the same item by different users. In order to improve the objectivity of online commodity evaluation, an online commodity evaluation method considering the inconsistency of user rating criteria is proposed to improve the objectivity of online commodity evaluation when different users' ratings on the same item are not comparable. Firstly, the incomplete scoring data is filled based on the similarity between users, and then the user's preference for online goods is calculated according to the filled user-commodity scoring matrix. Finally, based on the Schulze social selection function in group decision theory, commodity evaluation is obtained by constructing weighted digraph and calculating the strongest path. The evaluation result reflects the preference of the group, so that the consumers can make the correct purchase decision based on the evaluation results. At the same time, the method based on online commodity pairwise comparison improves the complexity of manipulating commodity evaluation results. The rationality and validity of the method are verified by theoretical analysis and experiments.
【作者单位】： 昆明理工大学信息工程与自动化学院;云南省计算机技术应用重点实验室;云南大学信息学院;
【基金】：国家自然科学基金项目(61462056,61472345,81560296,61462051)资助 云南省应用基础研究计划项目(2014FA028)资助
【分类号】：F724.6;TP393.092
[Abstract]:The inconsistency of the user rating standard leads to the incomparable rating of the same item by different users. In order to improve the objectivity of online commodity evaluation, an online commodity evaluation method considering the inconsistency of user rating criteria is proposed to improve the objectivity of online commodity evaluation when different users' ratings on the same item are not comparable. Firstly, the incomplete scoring data is filled based on the similarity between users, and then the user's preference for online goods is calculated according to the filled user-commodity scoring matrix. Finally, based on the Schulze social selection function in group decision theory, commodity evaluation is obtained by constructing weighted digraph and calculating the strongest path. The evaluation result reflects the preference of the group, so that the consumers can make the correct purchase decision based on the evaluation results. At the same time, the method based on online commodity pairwise comparison improves the complexity of manipulating commodity evaluation results. The rationality and validity of the method are verified by theoretical analysis and experiments.
【作者单位】： 昆明理工大学信息工程与自动化学院;云南省计算机技术应用重点实验室;云南大学信息学院;
【基金】：国家自然科学基金项目(61462056,61472345,81560296,61462051)资助 云南省应用基础研究计划项目(2014FA028)资助
【分类号】：F724.6;TP393.092
【相似文献】
相关期刊论文 前3条
1 尤梦丽;雷秀娟;;PPI网络聚类的评价方法的研究与应用[J];计算机科学;2013年12期
2 王t，
本文编号：2366363
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2366363.html
