Study On The Genetics Methods Based Personalized Recommdenation

E-commerce is a new method of trade that emerges and develops on the international market. On one hand the new commerce brings the enterprise new benefit, on the other hand it brings new challenges. How to attract customers and enhance customer loyalty is the key to the competition of enterprises. By collecting interests and favorites, personalized systems can help customers to find the items which they prefer to enhance customer loyalty. While the personalied recommendation model is the center of the recommender systems, therefore the personalized recommendation model have become a hot theme in E-commerce system.This thesis firstly introduces personalized recommendation system, association rules techniques in personalized recommendation systems, music genome project and so on. How to introduct the concept of gene used in biology to personalized recommendation systems to supply theoretical foundation for recommendation systems are mainly described.Secondly, directe against the problems that the current personalized recommendation models do not consider the influence of items'special attributes on user purchasing behavior, and the real reasons that users buy one item, and so on. After introducing the concept of gene used in biology, a model based on item gene is proposed, which found item gene preferred by users and inherited them to the chosen items. The model can identified users'purchasing motivation better, so it can improve the recommendation accuracy and customer satisfaction. Lastly, based on establishment of model which is based on item-gene, looking for the item-gene greater impact on user purchases, and in accordance with item-genes'hereditary produce collection of recommdation items. And by using of item-gene mutation, considering relationship between customers'purchases and time of customers'purchases, introducting customers tag to be new item-gene matains system gene pool, item gene pool and customer gene pool. Tracking changes in users'interests, produce collection of recommendation items which adapt to changes in user interest, and increase customer loyalty and satisfaction furtherly. In order to verify the feasibility of personalized recommendation model based on item-gene, achieve presented model in the form of Java code. Both achieve the desired results.