Telecom Business Trends Model

Data mining is an important tool in knowledge discovery. Its purpose is to find implicit in a lot of valuable information in the data model. Time series prediction use the important tool mining the time series potential within the rules and patterns to predict future trends for data mining. While the rapid development of Internet technology,database applications,field scope,scale and depth of expanding. Data mining is generating to obtain data from a large number of valid,novels,potentially useful and ultimately understandable patterns. Time series predict has been widely used in the business,financial,medical,meteorology,telecommunications and other industries. Time series trends analysis can be applied to predict the music trends and evaluate the music trends.With the development of telecom value-added services,global telecom operators are beginning to tap the real music market. From first introduced in Korea in 2002,has been made in the Asia Pacific CRBT brilliant success,and quickly spread to the global market. Thailand's largest telecom operator AIS operations in late 2003 after the introduction of ring tones,ring tones just 6 months surged to 2.5 million users. With the commencement of 3G network construction,and give full play ring tones with a large user base and business to carry out the advantage of easily and quickly download music to the whole extension of other digital music business has become more and more operators in the development of the digital music business Strategic development strategy. Ring tones no longer a simple value-added services,and will serve as a carrier and promotion of digital music channels to end users.According to different types that used for quantitative prediction of sales time series methods can be divided into two categories: time series analysis and regression correlation analysis. Time series analysis is use the correlation between variables and time,through the analysis of previous data to predict future trends. In the analysis of sales income,we all know how the sales revenue in accordance with the order of years or months down to observe trends. Time series analysis has become a sales representative prediction method. Regression and correlation analysis use the relationship that various things are direct or indirect to each other,the sales will also be a variable with the changes. When selling things exist to other relationship beside time,we can use regression and correlation analysis for sales forecasts.This paper predicts the time series from sales. According to previous sales and the use of the system within the built-in or user-defined model for sales predicting future ring tones available on the relevant sales and reach a good prediction.