Luxury E-commerce Recommendation Research And Application Of Collaborative Filtering Algorithm

In the recent years, the e-commerce of China with the rapid development of Internet technology is entering the maturity phase, we can say that e-commerce has entered our lives. This contrast, the luxury goods market in the field of e-commerce development is lagging behind, and the scale of Chinese luxury market is very asymmetric. According to recent reports, the total spending of Chinese mainland luxury market has reached over$100billion annually. China will substitute the America to become the biggest luxury goods market in the world within the next few years.In the e-commerce platform, the recommended system is an importantly technology. The recommended system uses filter technology to recommend different goods or content to the user may be interested in them, thereby enhancing the user experience and sales of the website. However, since luxury is a unique, rare, exotic features such as consumer goods, it is necessary to research for the recommended system of luxury e-commerce platform and to explore a recommendation algorithm for luxury.First, this paper introduces the definition of luxury goods and consumer behavior, and then describes the structure and classification of the recommendation system and the popular recommendation algorithm, then focus on various types of collaborative filtering algorithm in-depth research and analysis, and compared their respective features.The core content of this paper is to improve and optimize the Slope One algorithm. Slope One algorithm is an item-based collaborative filtering algorithm, the algorithm is simple and features make it efficient and better accuracy. Slope One has some advantages, but it ignores the problem of user similarity and computing range. Based on this, this paper presents a new algorithm that combining the user-based collaborative filtering algorithms and Slope One recommendation algorithm to improve the accuracy and efficiency. Experiments show that the new algorithm has better accuracy prediction and recommendation, and proves the value of this paper’s research.Finally, this paper utilizes the Apache Mahout to implement the luxury recommendation system that is based on the new algorithm, also study and discussion related technical in detail. The result of this paper provides a useful reference to recommended algorithm for luxury e-commerce platform.