Research On Intelligent Mining Technology For Massive Data Of Electronic Commerce

With the rapid development of computer and network technology, the Internet haspenetrated into all areas of people’s lives, it is changing the way people live in modern societyunwittingly. Now, making the content creation and sharing is becoming more and more easy.Vast amounts of network information have satisfied the demand of user’s information. However,too much information is definitely not a good thing, people gradually has entered into the era ofinformation overload from an era of scarcity. In this day and age, both information consumersand information producers had a lot of challenges: for information consumers, finding theinformation they are interested in from a lot of information is a very difficult thing; forinformation producers, making production information stand out is also a very difficult thing.As is known to all, in order to solve the problem of information overload, numerousscientists and engineers have proposed many excellent solutions. To sum up, divided into threemain stages, namely, navigation, searching and recommendations. Baidu and Google, that weusually use the search engines, are representative of the retrieval. We have to enter a cleardemand information to use them. The emergence of recommendation engines changed all that.With the development of the Internet and the support of national policy, the electroniccommerce have sprung up flourished. Under the condition of Internet, e-commerce sites tend tosell more goods than traditional retail stores. How to choose suitable for their goods from vastamounts of goods, this is the concern of the user; how to let the user choose their goods from thevast amounts of commodities, has become the bottleneck of e-commerce development.According to the above introduction, recommendation system is applicable to e-commerce sites.So E-commerce site recommendation system has become a hotspot of current research inacademia and industry.First of all, in this paper, we introduce the classification of the current recommendationalgorithm, and by the simple examples describes the principles of each recommendationalgorithm, joint amazon e-commerce site, illustrates the application of the recommendationsystem in the actual environment.The second, we developed an information collection extraction system. In this system, weuse Heritrix and JSoup to extract information. To test the performance of the system, we extractinformation from Tmall and Alibaba sites. The experimental results show that the system dowell.Once again, the item-based collaborative filtering algorithm and the user-basedcollaborative filtering algorithm face the problem of sparse matrix. The sparse matrix completionrequire too much storage space and too much time to calculate. In this paper, we haveinvestigated the use of Latent Factor Model for collective tag recommendation. Tagging hasemerged as a powerful mechanism that enables users to find, organize, and understand onlineentities. Tag recommendation has received considerable interest in recent years. Most work hasfocused on personalized tag recommendation, suggesting tags to the user bookmarking a newresource. This is often done using collaborative filtering, taking into account similarities between users, resources, and tags.Finally, we has carried on the evaluation of the algorithm, and obtained ideal results.