Study On Network Information Mining In E-Commerce

The current rapid development of Internet technology and E-commerce applications continued to deepen, Web-based business transactions has become a key factor for the development of E-commerce. Web provides a new trading platform, the resulting transaction speed and convenience of which makes traditional business activities across the geographical restrictions, shortening the distance between customers and e-business enterprises, promoting enterprises better understand their customers demand. The E-commerce platform, generated a lot of data, including transaction data, Web content data and the log data. As E-commerce involves large amount of data, the type and the model is complexity, and the application of the distribution and heterogeneity is special, which makes the analysis of the data very difficult. Network information mining is just a precise technology, which finds the law of these data. Network information mining bases on the traditional data mining or knowledge discovery on the database. On the application of E-commerce, Network information mining includes both large transaction database mining and data warehouse mining, also includes Web content mining and usage mining. So the network information mining technology for E-commerce has become increasingly important.However, on the one hand, the current E-commerce system design is to be carried out its own commercial logic, without too much consideration to the data analysis and mining, even using some intelligent search engines, personalized analysis and recommendation function, which is far from the characteristics and advantages of Network information mining; On the other hand, E-commerce data mining research just stay in the traditional large-scale transactions database mining, or just singling out the Web log. It is necessary to improve the mining model and the system architecture, make the process completely integrated into the entire E-commerce business processes, and combine the traditional large-scale transactions database and the Web-based content and log data. Only in this way can we better for the collection of adequate data, better for the implementation of data mining and analysis, can show better results and commence mining of deep-seated other applications.This paper, completely studies the basic functions of E-commerce processes, analysis of the data sources of E-commerce mining, and proposes a E-commerce-oriented network information mining model and the general logic framework of network information mining system. And then discusses the evaluation system of network information mining system. Finally, with a application example of tourism E-commerce personalized service system, this paper display the application significance of network information mining for E-commerce.Specific, this study includes the following aspects :First,the paper completely describes the basic functions of E-commerce processes. In-depth analysis of the data sources of E-commerce mining. Secondly, this paper points out the current basic theory and techniques of network information mining. And then proposes a E-commerce-oriented network information mining model, and focus on its E-commerce site, personalized recommendation services and the business intelligence applications.Third, this paper focus on the common logical structure of E-commerce network information mining, and the introduction of a J2EE-based network information mining system program.Fourth, this paper products business mining system evaluation and selection. From the aspects of commercial capabilities, algorithm, process, e-commerce capabilities, this paper redesign the evaluation system of network information mining system.Finally, the application examples of e-tourism personalized service system, display of the significance of network information mining in E-commerce. Furthermore, the general logic structure of E-commerce network information mining system in this paper can easily be used, and the evaluation system of which is a comprehensive. This study has important practical significance. From the perspective of the needs of mining, this paper completely describes of the basic functions of E-commerce processes and the resulting data sources. The E-commerce-oriented network information mining mode proposed by this paper involves not only path analysis, association rules, sequence mode, categories clustering methods, but also comprehensive account of the two different types of data sources: Web log data and transactions data. Theoretically, it is very common, essentially applies to the majority of the BtoC E-commerce enterprises.Finally, it should be noted that most of the examples in this paper is from a large tourism E-commerce enterprises the author once worked in. It has great practical significance. What shows in the last is the practical application of network information mining in the tourism e-commerce.