Group of Frank Neven

Prof. Dr. Frank Neven is vice-director of the UHasselt Data Science institute. His research involves the broad field of data management and data engineering. More information on his recent exploits can be found here. 

In our research so far, we investigated different measures for approximate functional dependencies. In particular, we achieved a formal comparison by introducing a novel framework that represents these measures in terms of Shannon and logical entropy. Additionally, we conducted a qualitative sensitivity analysis with respect to the structural properties of input relations and performed a quantitative evaluation of the effectiveness of AFD measures for ranking approximate functional dependencies on real-world data. Based on our findings, we provided clear recommendations on which AFD measures are most suitable for practical use.