Integrating text mining and genetic algorithm for subject selection

Abstract

Advances in science and technology have brought about a growing number of study areas and disciplines. This in turn, results in the increase of subjects or units being offered via modular programmes or courses in universities or institutes of higher educations. The modular method of selecting subjects for completion of a course can be likened to the process of selecting events to complete a timetable, albeit a possibly less constrained variation. This paper combines the possibility of text mining the web and applying timetabling strategies using genetic algorithm (GA) into the subject selection problem. It aims to provide a basis for a system that provides advisory selections for students of higher educations. A subject selection system integrating text mining and specific GA mechanisms is implemented. Experiments and test runs with randomly generated excess population of chromosomes indicated a fair degree of success in this implementation. © Springer Science+Business Media B.V. 2010.