Inconsistent data exists everywhere but not always evident. Undoubtedly, using inconsistent data may lead to poor decision making, expensive mistakes, communication chaos, etc. There is an increasing industry-driven demand for tools that can efficiently identify and reduce inconsistent data. One of important steps in developing such tools is to find ways of specifying the structure and semantics of data which will enable the detection of inconsistent data by efficiently checking against predefined specifications.
The project aims to first analyse patterns of inconsistencies occurring among data from the same or different applications, and then develop a formal framework and related algorithms which can capture inconsistent data based on such patterns. Fundamental problems associated with the framework (such as computational complexity, expressiveness and succinctness) should also be investigated if the project is undertaken as a PhD project.
The project requires basic knowledge of data mining, databases, and algorithms.