The Reaserch Of Data Quality Of Operation And Management Analysis System Of Tele-communication Based On ETL

With the increasing development of data warehouse technology, as the important method for supporting decision-making, this technology has been already accepted by more and more industries and department. Therefore, in Tele-Communication Industry, it is vital to build a Operation and Management analysis system.According to the experience of the author in the Operation and Management analysis system. This thesis is researched how to guarantee the data quality of the data in data warehouse of the Operation and Management analysis system before the data enter the data warehouse, exclude the potential data problem, for the good foundation of the excellent decision-making. It becomes the core problem when the data warehouse is being established, it is also the purpose of this thesis.Firstly, this thesis is designed a uniform ETL workflow model based on the frame of the Operation and Management analysis system data warehouse According to the comparison to the data character of other industries. The whole ETL processes subdivided into much more details. For each procedure, the system will record the process logs and process state, which will be written into ETL log.In addition, this thesis would be introduced the methods and steps of data quality throughout the whole ETL process in detail.Finally, this thesis will be introduced the practical application of this uniform ETL workflow model applied in a Operation and Management analysis system of a Telecom operator.Through the expanding and enlarging of the ETL model, designing a set of data quality check module based on the uniform ETL workflow model which can be applied in the Operation and Management analysis system data warehouse of the general Telecom industries, the module can classify the problems the system happens, and can also do the threshold setting and checking, what’s more, it can warn us through short message and E-mail when the data checking and ETL flow is abnormal, and it can support the rules dynamic increasing. Throughout a set of measures, it can effectively promote and control the data quality before the data enter the ODS level of Operation and Management analysis system data warehouse.