Research On ETL Scheduling Of Data Warehouse Based On Discrete Firefly Algorithm

With the widely use of such technology as data mining and data analysis, the value of information which contained in data has been concerned by more and more people. Data warehouse is a method to store and manage a large amount of data, and it can help enterprises to make usage of their data resource. ETL which refer to Extract-Transform-Load is a very important process of data warehouse construction and the efficiency of ETL process will determining the quality of data warehouse to a great extent.This paper studies the ETL scheduling of data warehouse. The paper summarizes the features of data warehouse and analyzes the ETL process of data warehouse. According to the background of ETL scheduling, the firefly algorithm is improved. A method based on discrete firefly algorithm is proposed, which can help to find a relatively least-cost dispatch plan under a multiple-processor condition. It is able to reduce the total processing time and improve the efficiency of the ETL process. At last, two ETL instances are taken to do the simulation test, which indicated the validity of the method.