Investigating Incident-Induced Congestion for Personalized Travel Demand Management