This paper attempts to model the process of activity scheduling conflict resolution using actual scheduling process data. The resolution of activity scheduling conflicts is a critical component of rule-based activity scheduling models. Many current scheduling models use an assumed priority for each activity type to estimate how activity conflicts will be resolved, but research has shown that these activity type-based priority assumptionsoften do not hold in actuality. Therefore, the conflict resolution data captured in the scheduling process survey was used to estimate and evaluate a number of conflict resolution models including a decision tree model and two discrete choice models. Both the conflict resolution decision treemodel and the discrete choice models showed a promising ability to predict the resolution strategies chosen based almost entirely on the attributesof the activities in conflict and characteristics of the surrounding schedule. These models present a useful advance in increasing the realism andaccuracy of rule-based activity scheduling models.
Abstract