The idea is to build a generic platform for implementing Evolutionary Algorithms upon libraries already developed like JGap, ECJ and others. These libraries are added following the eclipse plug-in architecture, in principle a plug-in that puts JGap 1.0 into unEvo will be written.

The main components of an Evolutionary Algorithms will be added like plug-ins, these components are:

• Libraries: including representation types, genetic operators and selectors.
• Stopping Criteria
• Statistics of two types:
  • Table Statistics
  • Graphic Statistics

The only thing an user must to code is the fitness function, it can be written in Jython, an scripting language over java, which is very easy to learn and appropriate for users who don't have programming knowledge.

unEvo is developed at the Intelligent Systems Research Lab (LISI), at the National University of Colombia.