A Mixed-mode E-commerce Is The Recommended Techniques

With the popularity and development of the Internet, E-commerce is increasingly in-tegrated into people's daily lives. But it is di?cult to find the products consumers interestbecause it is so tremendous. In this case, E-commerce recommendation system come intobeing and become important research contents of the E-commerce technology. In order tosolve some exist problems, some researches as follows are carried in this paper.Firstly, discovery of frequent patterns is one of the important parts of the E-commercetechnology. In this paper, an index table of pre-large items to their corresponding originaltransactions is proposed to find out the transactions need to be processed. Then we workout the frequent patterns by using compact FP-Tree and matrix based algorithm. The exper-imental evaluation shows that the proposed algorithm outperforms the pre-FUFP algorithm.Secondly, the collaborative filtering technology is one of the successful technologies.In this paper, the similarity of the predicted items was integrated into cosine similarity of theuser's ratings. Thus it could get more accurate similarity of two users and nearest neighborsof the target user. So more accurate predicted items could be recommended.Additionally, two methods which are products recommendation based on frequent pat-terns and collaborative filtering are integrated in this paper. The experimental evaluationshows that the integration could express their respective advantages and outperform the re-sults by using two separate recommended methods.Finally, the real data provided by MovieLens site is used to support the e?ectiveness ofthe algorithm proposed in this paper.As mentioned above, an E-commerce recommendation system integrating frequent pat-terns and collaborative filtering is achieved in this paper. It has improved the e?ectiveness ofsome key skills such as the achievement of new frequent patterns, the better recommendationresults by using similarity of the items and integrating two methods mentioned above.