VIKAMINE is a flexible environment for visual semi-automatic data mining written in pure Java.
It features several powerful visualization and mining methods, and can utilize background knowledge.

VIKAMINE 1.3 is now available! See the download page for the new version. This release fixes some minor bugs, and implements several GUI improvements concerning the navigation and visual mining techniques. It especially features a simplified graphical interface in order to further enhance the user experience, while still providing all the powerful features. Additionally, several new quality functions for the automatic mining methods and visual tools have been integrated for further improving the semi-automatic visual mining experience.

Download: here

Web Site: http://vikamine.sourceforge.net
Project: http://www.sourceforge.net/projects/vikamine/

VIKAMINE is a flexible environment for visual semi-automatic data mining written in pure Java.
It features several powerful visualization and mining methods, and can utilize background knowledge.
VIKAMINE 1.2 is now available! See the download page for the new version. This release fixes some minor bugs, and implements several GUI improvements concerning the navigation and visual mining techniques. It especially features a simplified graphical interface in order to further enhance the user experience, while still providing all the powerful features.

Also, the screencasts have been updated for the new version.

Download: here

Web Site: http://vikamine.sourceforge.net
Project: http://www.sourceforge.net/projects/vikamine

In theory the verification and validation of engineered knowledge bases is the most important step before releasing a system to the public. However, in practice it denotes a painful phase in the overall project since domain specialists need to review the built system in its entire functionality.
During a running project we learned that the verification/validation step can be significantly simplified by “analog methods” that help the experts to step away from the original system and to review the knowledge from a distinct perspective.
Read the rest of this entry »

We are currently doing some screencasts demonstrating the power of VIKAMINE. The first quick start screencast has been released, and is now available on the project site, or directly here.

Enjoy!

VIKAMINE is a flexible environment for visual analytics, data mining and business intelligence - implemented in pure Java. It features several powerful visualization and mining methods, and can utilize background knowledge.

VIKAMINE-1.01 is a maintenance release that fixes some minor bugs related to data import, and also contains some visual enhancements.

Download: here

Web Site: http://vikamine.sourceforge.net
Project: http://www.sourceforge.net/projects/vikamine

VIKAMINE is a flexible environment for visual semi-automatic data mining written in pure Java.
It features several powerful visualization and mining methods, and can utilize background knowledge.

The first major release, VIKAMINE 1.0, is now available! See the download page for the new version. This release fixes some minor bugs concerning the data import, and features several GUI improvements concerning the navigation and visual mining techniques.

Download: here

Web Site: http://vikamine.sourceforge.net
Project: http://www.sourceforge.net/projects/vikamine

The yearly workshop series is taking place in Würzburg this year and hosts workshops of the german special interest groups “knowledge management”, “knowledge discovery & machine learning”, “information retrieval”, and “adaptivity and user modeling”. 

The deadlines have been extended or will be extended in the next days, so check out if you can/want to submit a paper. Go to the official website.

VIKAMINE is a flexible environment for visual semi-automatic data mining written in pure Java. It features several powerful visualization and mining methods, and can utilize background knowledge.

The new VIKAMINE release is now available! See the download page below for version VIKAMINE 1.0RC4. This release fixes some minor bugs on the road to version 1.0.

Download: here

Web Site: http://vikamine.sourceforge.net
Project: http://www.sourceforge.net/projects/vikamine

The versatile data mining environment VIKAMINE moved to SourceForge. Thus, the development and future releases will be eased by the great services provided by SourceForge.

Since 2003 VIKAMINE has been developed by the Department of Artificial Intelligence and Applied Computer Science at the University of Wuerzburg. The released files already moved to SourceForge servers. Mailing lists and tracking system have already been set up, and SVN will follow shortly.

http://vikamine.sourceforge.net

Two days ago it was “Girls’Day”! Here are some information from the official press text:

“On the 24th of April 2008 technical enterprises, enterprises with technical departments and technical training facilities, universities and research centres are invited to organise an open day for girls - the Girls’Day.”

Some school girls visited us at the University Würzburg to see how it would be to become a computer science student. Well, they did not only listen to professor’s and student’s talks but also were invited to participate in practical projects in small groups. My colleague Jochen showed one group how to build “smart wikis” using our semantic knowledge wiki KnowWE. Although he proposed a lot of interesting alternatives, they (as many others before) chose to build a “movie recommendation wiki” - probably the classic one out of all recommenders…