基于推荐完整性的电子商务推荐系统架构的设计

Design of recommendation system architecture based on integrity

针对个性化推荐完整性欠佳的问题,设计一种基于第三方的电子商务推荐架构。使用聚合内容获取信息源,运用多代理实现推荐流程,借助面向对象存储技术在第三方平台扩充备选项集。通过对上述技术方法的组合应用,使推荐的完整性得以提升,克服了传统推荐系统忽略用户站外行为影响推荐完整性的缺陷。

This paper designed one kind of architecture of recommendation system for e-commerce based on third-party,which aimed to conquer the problem of neglecting integrity of personalized recommendation. The architecture gathered information by RSS,realized recommending process by multi-agent system（ MAS） ,expanded option sets by object-based storage within the platform of third party. Enhanced the integrity of recommendation by the collaborative application of the above methods. This architecture conquered the problem of neglecting users＇outside preference which existing in popular recommendation systems.