Development of Recommendation System of Place of Employment for Students with Decision Rules in Rough Set using the Condition Relaxation based on User’s Aptitude

Abstract

Conditional search system in employment supporting website is well used in Japan, if college or university students search place of employment. However, this system is not possible to respond flexibly to user’s undefined preference, and recommends only a few place of employment with strict conditions setup.

This paper proposes Development of the recommendation system of Place of Employment for Students with Decision rules in Rough Set using the condition relaxation based on user’s aptitude. First, this proposal system changes the attribute value of target place of employment into qualitative values. Next the system searches the place of employment which corresponds with the target user’s absolute conditions. When the number of search is not enough, the system eases conditions and searches additional places of employment, which are estimated by standardization Euclid distance, and which are selected by aptitude judgment. Finally, this system extracts decision rules of rough set based on the samples' judgement, and evaluates and shows places of employment, which are sorted in the descending order of the evaluation value. Through practical experiment, the proposal system recommended Place of Employment for Students more effectively than the usual Condition search.