Combining linear programming and clustering techniques for the classification of research centers

Article type: Research Article

Authors: Tagarelli, Andrea | Trubitsyna, Irina | Greco, Sergio

Affiliations: DEIS – University of Calabria, 87030 Rende, Italy E‐mail: {tagarelli, irina, greco}@si.deis.unical.it

Note: [] Corresponding author.

Note: [] S. Greco is partially supported by ICAR‐CNR and Exeura S.r.l.

Abstract: We propose a novel approach that combines data mining and linear programming techniques for classifying organizational units, such as research centers. We show how our proposal of clustering organizational units based on both efficiency and input/output parameters turns out to be effective in identifying groups of similar organizational units. We also propose the replacement of an expensive efficiency measurement, based on the solution of linear programs, with a simple but more efficient formula to be exploited in the clustering process. Preliminary experimental results, obtained from an analysis of research centers in the agro‐food sector, show the effectiveness of our approach.

Keywords: Clustering, data envelopment analysis, efficiency measures

Journal: AI Communications, vol. 17, no. 3, pp. 111-122, 2004