Software

We would like to announce the 2.0 release of CLUTO, a software package for clustering low and high-dimensional datasets. CLUTO is well-suited for clustering datasets arising in many diverse application areas including information retrieval, customer purchasing transactions, web, GIS, science, and biology. CLUTO provides a wide-range of partitional and agglomerative clustering algorithms, and a variety of tools for analyzing and visualizing the discovered clusters.

CLUTO Release 2.0 contains a number of major additions over its earlier release. Some highlights of the new release are the following:

• CLUTO provides a new class of clustering algorithms based on the graph-partitioning paradigm. These algorithms are well-suited for finding non-globular clusters. The graph-partitioning is based on the highly efficient multilevel algorithms implemented in METIS and hMETIS.

• All of CLUTO's clustering algorithms have been extended to operate on a user supplied object-to-object similarity matrix. As a result, CLUTO can now be used for clustering arbitrary datasets (eg., protein sequences, 3D structures, etc.).

• CLUTO can compute a clustering solution by combining both partitional and agglomerative approaches. This framework was motivated by the CHAMELEON clustering algorithm and can be used to find non-globular clusters.

• CLUTO's agglomerative algorithm has been extended to include the traditional single-link, complete-link, and group-average schemes.

• CLUTO now supports Euclidean-distance based similarity measures.

Obtaining CLUTO

CLUTO is freely distributed. Information on how to get the package is available on WWW at:

URL: http://www.cs.umn.edu/~karypis/cluto

Contact Information

CLUTO has been written by George Karypis, at the Computer Science Department of the University of Minnesota. If you have any questions or problems obtaining CLUTO, send email to karypis@cs.umn.edu.

CLUTO is copyrighted by the Regents of the University of Minnesota.