Today many data warehousing systems are data rich, but information poor. Extracting useful information from an ocean of data to support administrative, policy, and instructional decisions becomes a major challenge to both database designers and measurement specialists. This paper focuses on the development of a data processing system that integrates multiple types of analyses to support the maintenance and further development of a large-scale online assessment efficiently. An automation package has been developed that involves three components of data processing using a Perl script as the master program: (1) data parsing through Perl; (2) data analysis with Iteman, Winsteps, Bilog/Bilog MG, LISREL, and the Statistical Analysis System (SAS); (3) data presentation through Xlisp-Stat/DataDesk, Web page generation. With this online and automated system, content experts could examine the psychometric attributes of test items in a more efficient manner. (Contains 6 figures and 12 references.) (Author/SLD)