The WEKA Environment for Knowledge Analysis has emerged as the mainstream open-source data mining software, maintained by Waikato University of Waikato.
WEKA features a multitiude of algorithm implementations that support a variety of general-purpose data mining tasks such as:

  • classification
  • clustering
  • regression
  • feature selection

In contrast to other data mining environments such as R, WEKA features advanced user interfaces that allow data mining experts to develop and test their algorithms interactively.
Visualization has been a major part of WEKA, and is often used by data mining experts to analyze their the characteristics of their data features.
This projects aims to empower WEKA with advanced interactive visualization techniques.
Besides the current possibilities to analyze data features, the projects aims to

  • provide intuitive visualizations of the results of data mining algorithms
  • enable advanced performance analysis, in terms of result accuracy, to allow investigating possible sources of erroneous results
  • provide interactive means to adjust or extend the algorithms, and to test the improvements

Members: Bilal Alsallakh, Ahmad Bisher Tarakji, Emmanuelle Beauxis-Aussalet