The Research And Application Of Data Mart In The Telecommunication Business Analysis

With the goal of the telecom operators from increasing the number of customers to improve customer quality, take customers as the center, provide them diversification, multi-level, personalized service solutions, the requirement of data mining and analysis is also getting higher and higher. Build enterprise data marts can help telecom operators to analyse customer behaviour, identifying customer consumption characteristics, auxiliary them for effective marketing and customer service. Making full use of data information related to marketing and management decision-making can provide telecom operators seasonable, effective, scientific decision-making in the customer relationship management process.Now the company business analysis still with artificial statistics, and data mining technology has not been reasonable use, the process of statistical analysis is time-consuming and laborious, and lack of data accuracy and consistency. For these reasons, the thesis put forward to data mining and analysis for the purpose, and building data marts for the method to support the urgent data requirements in business analysis, the research content mainly includes the following three aspects:i. The key feature of data marts are designed and researched, built the model structure of data mart, and clear the subject domain of data mart according to the needs of business analysis, designed the data interface, and integrated the source of different data.ii. Designed the process of the ETL detailed, determined the standard and strategy of data extraction and data loading, optimized and improved the data cleaning and the SQL statements of data transformation, built the fact table commonly used in business analysis, does a good job on the management of metadata, and optimized the system performance.iii. Developed the front display interface of data mart based on JSP technology.The interface using WEB browsing, combined with charts and other data display forms, make the data more easily to be understood, excellent support business analysis.