Abstract:

Expert systems have evolved as very robust and customizable tools for solving complex problems that typically require the knowledge of domain experts and a human's decision-making skills. They are especially useful for automating the decision making process for problems that would take inordinate amounts of time to solve by a human. Evaluating the data quality of an organization in multiple dimensions and assessing the confidence and reliability of uncertain (fuzzy) data are two such problem. In this project, I will develop a framework for evaluating the quality of uncertain data sets using a fuzzy expert system that will be built to evaluate the data quality of an organization that works with GPS positioning data in terms of security, integrity, reliability, and accuracy. Simulated GPS positioning data will be used for evaluation because the accuracy of the data is uncertain and requires fuzzy processing to effectively evaluate. The expert system will provide qualitative and quantitative feedback about the data quality assessment of the organization, as well as recommendations to remedy deficiencies.

All documents (the proposal, project report, source code, etc.) can be found on the Documents page.

Committee:

Chair: Dr. Leon Reznik

Reader: Dr. Richard Zanibbi

Observer: Dr. Roxanne Canosa