Specific information about travelers' activities is often vital for transportation planners. For example, in modeling destination choice should only those attractions where a desired activity can be pursued be included? Destination choice modeling may improve when more specific activity information is available, especially for leisure trips. Destinations for leisure trips can be places to visit people, cinemas, and restaurants. In surveys, however, respondents are often asked to categorize their trips into generic purposes such as work, shopping, or simply leisure, without specifying the leisure activity. Hence, there is often a gap between trip-purpose information elicited from respondents and specific activity information, which would enhance destination modeling. A model is presented that bridges this gap with respect to leisure activity engagement. Common experience provides an idea of who is likely to conduct a particular leisure activity and at what time. The approach taken here was to quantify this likelihood by evaluating a data set of more than 20,000 leisure trips that provided specific activity information. This provided the calibration for the model IRMA, which simulates a specific activity for each trip based on trip-context data. IRMA was applied to model activities for leisure trips in a data set from Tuebingen, Germany. The result of the application of IRMA was consistent with common experience and provided vital activity information that helped to improve destination choice modeling. The approach to infer activity information from available trip data has proven useful in bridging the gap between information that is needed and information that is available.
Abstract