Common lecturers timetabling among departments based on funnel-shape clustering algorithm

摘要

The university course timetabling problem is considered one of the NP-problems which should be performed for each semester repeatedly, and it is an exhausting and time consuming task. The main technique of the proposed approach is focused on extension and scalability of common lecturers timetabling process across different departments of a university. In this paper the clustering algorithms, including K-means and Fuzzy c-means, are used to schedule common lecturers between departments considering the constraints and their priorities offered, for the first time. For this purpose, a new clustering algorithm named funnel-shape clustering is proposed to schedule common lecturers between departments based on their constraints proposed. The main goals of this paper are to improve the satisfaction of common lecturers across departments and minimize the loss of resources within departments. In this method, all the departments perform their scheduling process locally; After that the clustering agent is applied to cluster common lecturers across departments by using the proposed funnel clustering algorithm and then the traversing agent is used to find the useless resources across departments. Following the clustering and traversing processes, mapping is performed based on the common lecturers’ constraints in the excess resources in order to reach the problem goals. The applied dataset based on the real world requirements is across various departments of Islamic Azad University, Ahar Branch and the results’ success would be based on uniform distribution and allocation of common lecturers on useless resources across different departments.