The Research And Application With Web Mining In E-Commerce Recommendation System

Since the 20th century, the network has deepens to our homes and spurs the new business technology——the development of e-commerce. The system for users to guide users to undertake choosing, convenient purchase needed goods. But because of commodity types of increasing, users often difficult to from a huge amount of accurate find itself catalogue needed goods. The recommendation system is to solve this problem and of generation, it can directly tracking user behavior, exploring the needs of users, liking shopping malls as sales staff to recommend commodities to the user, to help them find needed goods, thus promote the trade reached. The increasingly fierce competition in goods, goods under the environment of recommendation system can help distributors successful attract clients, reduce customer churn, promote enterprise's sales force and competitiveness.The recommendation system by this one characteristic, attend a lot of researchers, this system development and utilization will push forward the development of e-commerce field mass.The E-Commerce using the recommended system's knowledge about discovering technology can make personalized recommendations on products and services to different users. The application of recommender systems will definitely redound to enhance E-Commerce competitiveness in the industry.At present, although E-commerce recommendation system in the goods under the research of scholars have made abundant research achievements, but the E-commerce system, the scope of expanding the scale of promotes gradually, the commodity recommend system also face more new challenges. In view of the commodity recommendation system of main facing challenges, this paper focused on the E-commerce recommendation system is the following three aspects of exploration and research.1. Described some basic concepts and basic knowledge of recommendation systems; researched the date preprocessing of the recommended work flow in E-Commerce recommendation system, designed and realized the date preprocessing module.2. Current clustering algorithms most do not have the ability to dynamically adjust as users browse behavior change. In this paper, aiNet clustering algorithm based on incremental clustering algorithm for the current problems of the characteristics of the artificial immune system and ant colony clustering algorithm for incremental combining the idea proposed incremental clustering based on artificial immune algorithm.Meanwhile, this paper proposed algorithm in the experiment. Experimental results show that the increment based on artificial immune clustering algorithm error and the average classes in terms of algorithm execution time are reflected in good performance. 3. Based on artificial immune incremental clustering algorithm, designed the Web-based E-Commerce recommendation system model.