一种新的基于cookie的互联网个性化推荐系统设计

A New Design of Internet Personalized Recommendation System Based on Cookie

将cookie技术与模式识别算法相结合,基于对用户历史使用习惯的数据挖掘,实现了互联网的个性化推荐等功能。提出的特征映射与聚类相结合的模式识别算法解决了SPADE算法不能反映类信息的问题与大量数据聚类时的效率低下,同时提出了与本算法相配合的cookie设计,从而完成了个性化推荐系统的整体设计。对于理解互联网用户行为,并且以此为基础提升用户体验有着相当重要的作用。

This paper combines cookie technology with pattern recognition algorithm, based on mining for using history data, to realize internet functions such as personalized recommendation. This algorithm applies feature mapping together with clustering theory to solve the problem that SPADE algorithm could not reflect the cluster information and the algorithm efficiency is low deal- ing massive data. By also putting forward the cookie design which is matched with this algorithm, this paper finishes the whole de-sign of personalized recommendation system. This paper plays an important role in both understanding behaviors of internet users and improving user experience.