Research Of The Integration Analysis And BI Front-end Display Based On Marketing Data Of The Power Industry

After more investment from the power industry, it has got some achievement in the database construction and a variety of applications software development. However, most were uniquely developed for specific applications, they were poor for data sharing and no longer meeting the company’s current operations and decision-making. So, BI (Business Intelligence) which is a type of technology to solve this problem was recommended into the power industry. In this paper, We proposed integration analysis and display method and designed data integration and BI front-end display model of the power industry marketing data based on studied the current domestic and international ETL, data warehouse and BI front-end theory and techniques.The paper encircled the integration analysis and BI front-end display of marketing data in the power industry. Firstly, we introduced the ODS (Operating Data Store), DW (Data Warehouse) technology concept and design model of data warehouse, data mart, OLAP. And we suggested that the present front-end BI display encountered bottlenecks as well as the tools were used in this paper. Secondly, we designed ETL of heterogeneous data and data acquisition and data sharing methods; Finally, we built some kinds of BI display models. For example, KPI, fixed reporting, multidimensional analysis, real-time queries.We divided into three modules in the function.1, ODS module is including the ODS data model, analysis of data items and integration of heterogeneous data through Informatica; 2, DW module is including the design and implementation of dimension and fact table; 3, Data Mart module is including the power, electricity prices, sales,95598and customer relationship analysis. In the BI front-end display, it achieved a unified data view, it provided a uniform way to take the data to improve the timeliness of statistical analysis, so that business analysts could concentrate data analysis more.The paper is extreme significance. In the first place, it was not only the effective integration of heterogeneous data of the large-capacity, multi-system but also established data warehouse which differed from previous. In the next place, front-end display was not only beautiful but also supported from simple queries to drill-down analysis. What matter most, the techniques and methods are high versatility, a little change can be applied to other areas.