The Research and Application of Purchase Amount Based User-KNN Algorithm on Cloud Platform in Coal Sales System

摘要

With the highly development of internet e-commerce, coal shopping on-line is becoming a popular trend. With the increase of purchase data, every big coal industries hope to recommend the products which the customers will be interested with through analyzing the purchase history data. In coal system, purchase data based near neighbor algorithm is applied widely. However, with the increasing of huge big scale consuming records, traditional algorithm could complete recommendation work real-time and effectively. In this paper, focusing on huge scale coal consuming records, we propose cloud platform based user neighbor algorithm, and this algorithm is based on MapReduce distribute framework. It can complete the recommendation work distributed. Through the experimental results, we prove that the algorithm we propose has good efficiency and high scalability.