Decision Tree Based Routine Generation (DRG) Algorithm: A Data Mining Advancement to Generate Academic Routine and Exam-time Tabling for Open Credit System

In this paper we propose and analyze techniques for academic routine and exam time table generation for open credit system. The contributions of this paper are multi-folds. Firstly, a technique namely Decision tree based Routine Generation (DRG) algorithm is proposed to generate an academic routine. Secondly, based on the DRG concept, Exam-time Tabling algorithm (ETA) is developed to implement conflict free exam-time schedule. In open credit course registration system any student may choose any course in any semester after completion of pre-requisite course(s). This makes the research more challenging and complex to accomplish. Academic routine and exam timetable generation are in general NP-Hard problems, i.e., no algorithm has been developed to solve it in reasonable (polynomial) amount of time. Different methods based on heuristics are developed to generate good time-table. In this research we developed heuristic based strategies that generate an efficient academic routine and exam time-table for a university that follow open credit system. OLAP representation helps to classify the courses along with the proposed algorithm to eliminate some constraints. Daybased pattern, minimum manhattan distance between courses of same teacher; minimum conflicted course distribution has been stage-managed to classify the courses. Our ETA algorithm is based on decision tree and sequential