Collaborative Inference Detection

Abstract

Access control mechanisms are commonly used to provide control over who may access sensitive information. However, malicious users can exploit the correlation among the data and infer sensitive information from a series of seemingly innocuous data access. In this thesis, I proposed a detection system that utilizes both the users current query and past query log to determine if the current query requests or results can infer sensitive information. Detection system is being extended to the cases of multiple collaborative users based on the query history of suspected database users and considers access levels of the users to infer speci c sensitive information. This thesis attempts a thorough coverage of the advancements in methods of inference detection by considering both schema and data level of the database.. There are a number of methods each with its own advantages and disadvantages. Design models contribute to the mitigation of inference occurrence. Functionality of the design models can be added to database front-end. Inference detection system has ability to completely detect inference with having high data availability. End of the thesis, I'll discuss the areas to be developed, in order to enhance the suggested interference detection system to be a sound inference detection system.