Research on Collaborative Filtering Recommendation Algorithm Based on Mahout and User Model

Abstract

Based on the research status of recommendation system, the traditional collaborative filtering algorithm is explored. By establishing the concept of training sample and Mahout framework, a collaborative filtering algorithm based on Mahout and extracted user model can be proposed. Using this algorithm in the system of recommendation aims to solve the problem of transforming the traditional recommendation problem into the classification problem in machine learning. In this way, it can provide personalized recommendations for users. The analysis of the experimental results of the collected some data sets suggests that the algorithm based on user model and machine learning is effective and optimal, at the same time, the efficiency of the algorithm under Mahout framework is also verified.