Research On Users’ Repurchase Behavior Based On Data Mining And Data Warehouse

Involving the fiercely competing and diversified market economy,continue to develop measures to maintain business evergreen become the key of the survival of the enterprise,the right decisions will lead the enterprise to the good development.Data plays an increasingly important role in business decision making,from the beginning spreadsheet to today’s data warehouse,the era of big data encourages the enterprises to pay attention to and make use of the data and make accurate decision service for enterprise business.The data is of immeasurable value, mastering the value in a certain sense means occupying the market share, turning the potential value into real business revenue.The development of data warehouse technology meets the need of big data storage,the development of data mining technology provides convenience for the enterprises to gain value from the data,the combination of the two is to provide decision support for more enterprises.With the increase of video website paid content, how to improve the user’s conversion rate of pay, to guide the user through the consumption data, has become the key to winning the video website users.The website users need to pay for single purchase and package purchase, whether to buy or buy a single package, to a certain extent reflect the user’s consumption potential.Excellent recommendation system will provide the video content for users willing to pay, through the analysis of user behavior data mining, guide the user consumption.With the increase of the number of video sites, users are also facing more and more choices.Video sites need to face the problem of churn,predict the user repurchase behavior in advance to make corresponding marketing strategy, it has great effect to cultivate loyal customers, for the development of the site has great significance.In the field of smart TV, provide good service can improve customer loyalty and TV brand reputation.Based on the above problems, this paper focuses on the customer repurchase behavior research, proposed repurchase behavior prediction method based on data warehouse and data mining technology.First is smart TV users’log analysis, determine the design of data warehouse, and then refer to the RFM model commonly used in customer relationship management to obtain the characteristic index, at last, use the decision tree algorithm on the user’s purchase behavior is predicted.The main contributions of this dissertation are:The first:summarize the research background and research status of the data warehouse and data mining technology, aiming at the current problems and put forward solutions.Second:To summarize and analyze the data warehouse and data mining technology, focuses on the building process of data warehouse, data mining and how to support the work of the.Third:the RFM model commonly used in customer relationship management, generate the user’s consumption index, and based on these indices, using C5.0 decision tree algorithm to predict the user’s repurchase behavior.