InvolvD is an ANR-financed (ANR-20-CE23-0023) project bringing together researchers of the GREYC laboratory of the University of Caen Normandie, the LIFO laboratory of the University of Orleans, the LABRI laboratory of the University of Bordeaux, and the CERMN research center, also situated in Caen.
InvolvD stands for “Interactive constraint elicitation for unsupervised and semi-supervised data mining” and is situated at the intersection of structured pattern mining and clustering, with a strong focus of bringing the user into the data mining loop. The two data mining tasks interleave in that
- structured patterns are used to describe structured data, i.e. molecules, to allow the use of a wide range of clustering approaches, and
- clustering provides additional information that can be exploited for the pattern mining task, moving from the unsupervised setting, where only the data itself is available, to a semi-supervised one, where we have information about the relationship among data.
The user exploits her or his expert knowledge to add additional feedback to assess the interestingness or plausibility of patterns or clusters.
Contact: Albrecht Zimmermann