Size estimation for data warehouse systems by using function point analysis

Abstract

Software size estimation has an important role in software development. Estimation is used not only for the input of project planning, but also for the management of the project. Function Point Analysis is a popular estimation method developed by Albrecht and Gaffney at IBM, which is now managed by the IFPUG (International Function Point User Group). It enables the early stage estimation in the software life development cycle. However, adopting Function Point Analysis to Data Warehouse system that has increasing demands from market is not adequate because of the characteristics of Data Warehouse: Data centric processing, several ETL (Data Extraction, Transformation and Transportation, Loading) steps to implement data warehouse, domain specific commercial tool such as OLAP (On-Line Analysis Processing) and reporting tools. In this paper, procedure and counting rules for applying Function Point Analysis to data warehouse application are proposed. In an empirical study, seven data warehouse applications from industry projects were analyzed using the proposed method. Based on the results, it was possible to establish a more accurate function point analysis method for data warehouse applications.