SIS Seminars

To be effective in the information age, managers should have the ability to understand, diagnose and solve data quality problems that arise in their firms. Clearly having quality data in the right form at the right time is the sine-qua-non for organizational success. The primary focus of this research seminar is to present several analytical frameworks, models and procedures for enhancing data quality in organizational contexts. These models together provide a framework for determining optimal resource allocation for maintenance of data integrity, to attain significant improvements in the information product manufacturing systems and for enhancement of data quality in data warehouses.

Maintenance options for data sets differ in effectiveness and cost. Further maintenance becomes complex since the error rates and impact of those errors on the applications are not accurately known. This data integrity maintenance problem can be analytically modeled as an integer program, which is solved using special heuristics to obtain solutions that account for the inherent uncertainty while maintaining specified levels of data quality.

Modeling the information manufacturing system along the dimensions of timeliness, quality, cost and value is critical to ensure that the information product is fit for use by the end consumer. The framework builds on a set of ideas, concepts, models and procedures that form the building blocks of the information manufacturing system. Using these building blocks a quantitative methodology is proposed for determining the information product attribute values and to study options for improving the information manufacturing system.

With rapid growth in the deployment and use of data warehouses firms are faced with the problem of enhancing the quality of the data that resides in them. Here an analytical framework is provided for thinking comprehensively and systematically about data quality enhancement. Specifically, a methodology based on an integer programming model is presented that systematically incorporates factors such as the current level of data quality, the levels of quality needed by the relevant decision processes and the potential benefits of projects designed to enhance data quality.