用户在线购买预测:一种基于用户操作序列和选择模型的方法

发布时间：2018-03-06 00:00

本文选题：在线购买预测　切入点：选择模型　出处：《计算机研究与发展》2016年08期 　论文类型：期刊论文

【摘要】：电商网站的兴起与用户在线购物习惯的形成,带来了海量的在线消费行为数据.如何从这些行为数据(如点击数据)中建模用户对相似产品的比较和选择过程,进而准确预测用户的兴趣偏好和购买行为,对于提高产品的购买转化率具有重要意义.针对这一问题,提出了基于用户行为序列数据和选择模型的在线购买预测解决方案.具体而言,1)使用行为序列效用函数估计用户在购买周期(session)中的最佳替代商品,然后对购买商品和最佳替代商品建立基于潜在因子的选择模型(latent factor based choice model,LF-CM),从而得到用户的购买偏好,实现对用户购买行为的预测.更进一步,为了充分地利用用户在每个购买周期的所有选择和比较信息,提高预测精度;2)提出了一种可以作用于购买周期内所有商品的排序学习模型(latent factor and sequence based choice model,LFS-CM),它通过融合潜在因子和行为序列的效用函数,提高了购买预测的精度;3)使用大规模真实数据集在分布式环境下进行了实验,并与参照算法进行了对比,证实了所提出的2个方法在用户在线购买预测上的有效性.
[Abstract]:The rise of e-commerce websites and the formation of online shopping habits of users have brought massive online consumer behavior data. How to model the process of users' comparison and selection of similar products from these behavior data (such as click data). Therefore, it is very important to predict the interest preference and purchase behavior of users, which is of great significance to improve the conversion rate of purchase. An online purchasing prediction solution based on user behavior sequence data and selection model is proposed. Then, a latent factor based choice model based on latent factor based choice is established for purchasing goods and the best alternative goods, so as to obtain the user's purchase preference and realize the prediction of the user's purchase behavior. To take full advantage of all user selection and comparison information for each purchase cycle, This paper proposes a sort learning model that can be applied to all items in the purchase cycle. It can be used to combine the utility functions of potential factors and behavior sequences. The accuracy of purchasing prediction is improved. (3) A large scale real data set is used to carry out experiments in a distributed environment, and compared with the reference algorithm, the effectiveness of the two methods in online purchase prediction is verified.
【作者单位】：中国科学技术大学计算机学院;
【基金】：国家杰出青年科学基金项目(61325010) 国家自然科学基金项目(61403358) 科技惠民计划项目(2013GS340302) 青年创新促进会会员专项基金项目(2014299) 多媒体计算与通信教育部-微软重点实验室基金项目~~
【分类号】：TP391.3

【相似文献】