Solving the preference-based conference scheduling problem through domain transformation approach

• Main Authors: Abdul Rahim, Siti Khatijah Nor, Jaafar, Amir Hamzah, Bargiela, Andrzej, Zulkipli, Faridah
• Format: Conference or Workshop Item
• Language: English
• Published: School of Computing, UUM College of Arts and Sciences 2017
• Online Access: http://psasir.upm.edu.my/id/eprint/64456/1/PID168-199-207e.pdf; http://psasir.upm.edu.my/id/eprint/64456/

Conference scheduling can be quite a simple and straightforward problem if the number of papers to be scheduled is small. However, the problem can be very challenging and complex if the number of papers is large and various additional constraints need to be satisfied. Conference scheduling with regard to satisfying participants’ preferences can be under-stood as to generate schedule to minimize the clashes between slots or sessions that participants are interested to attend. Motivated by the current re-search trend in maximizing participants’ satisfactions, the study looks at the possibility of scheduling papers to sessions without any conflict by considering preferences by participants. In this research, preferences refer to the papers chosen by participants that they would like to attend its’ presentations sessions. Domain Transformation Approach (DTA), which has produced very encouraging results in our previous works, is used in this study to solve preference-based conference scheduling problem. The purpose of utilizing the method is to test the generality and universality of the approach in producing feasible schedule, given a different scheduling problem. The results obtained confirm that DTA efficiently generated feasible schedule which satisfies hard constraints and also fulfills all the preferences. With the generated schedule, all participants are able to attend their sessions of inter-est. In the future work, additional constraints will be taken into account in optimizing the schedules, for example balancing the number of papers as-signed to each timeslot, and minimizing assignment of presenters to different timeslots. Other datasets could also be tested in order to test the generality of the proposed approach.