Thomas Karagiannis Home Page


version 0.1b

The SELFIS project is supported by the NSF CAREER grant ANIR 9985195, and DARPA award NMS N660001-00-1-8936, and NSF grant IDM IIS-0208950.

SELFIS (SELF-similarity analysIS) is a java-based software tool for self-similarity and long-range dependence analysis. It is freely distributed as a service to the community. For more information please refer to the corresponding papers and read the short tutorial of SELFIS.

Currently, it incorporates various long-range dependence estimation methodologies and the bucket shuffling methodology for long-range dependence detection.
Future releases will include time-series processing algorithms and transforms, such as wavelets, Fourier transform, stationarity tests and smoothing algorithms. In addition, the next release will include FGN and FARIMA processes generators to simulate long-range dependence in time-series.

Supported Features:

Note that this is still an early BETA release. Please inform us of any bugs that we missed at

If you publish any material using the SELFIS tool, then please send the citation information for your article to
Please use the following references for the SELFIS tool:

Thomas Karagiannis, Michalis Faloutsos, Mart Molle. A User-Friendly Self-Similarity Analysis Tool.
Special Section on Tools and Technologies for Networking Research and Education, ACM SIGCOMM Computer Communication Review, 2003(to appear).

Thomas Karagiannis, Michalis Faloutsos. SELFIS: A Tool For Self-Similarity and Long-Range Dependence Analysis.
1st Workshop on Fractals and Self-Similarity in Data Mining: Issues and Approaches (in KDD)
Edmonton, Canada, July 23, 2002.

You will need the Java Runtime Environment (JRE) version 1.4.1 or higher. You can download JRE from

Please complete the form below before downloading SELFIS.

Last Name:
First Name: