Modeling social interactions between individuals for joint activity scheduling

Joint social activities, in particular those outside households, are currently ignored or modeled very simply in transport models, despite these sorts of activities contributing to a significant amount of travel. We describe an experimental model of social activities, in which individuals negotiate about the type, purpose, location, and days of activities. After participating in activities, individuals learn about new locations and acquaintances. Using concepts from the activity-based modeling and social networks fields, a prototype model was created using Python incorporating utility-based agents who used a protocol to communicate with each other about potential activities in order to negotiate a suitable day and location. It can be shown that agents with a large number of acquaintances participated in more activities. Pairs of agents with high similarity values, based on age and gender, also socialized with each other more often. Future work involves further development and validation and eventual incorporation into activity-based models.