Table 1.
Term | Meaning |
---|---|
Activity Types for
Fuzzy Classification |
|
Walking* | Walking activities are those when an individual walks outdoors. |
Staying | Staying activities are those performed when an
individual stays at or around a location, such as shopping, staying at home, indoor walking, and dining out. |
Other Transportation | Other transportation activities include other
motorized or non- motorized transportation such as driving a car, bus riding, and cycling. |
Baseline Activities | Baseline activities are the group of
activities that 1) were recorded by the three subjects in journals and confirmed/corrected by the subsequent GIS validation and 2) that were not recorded by the journals but identified by the subject during the GIS validation. Information of baseline activities serves as the reference for evaluating the performance of the two methods developed in this study. |
Membership Function | In fuzzy logic methods, membership functions
are defined as the function that projects an input crispy value into a membership degree that ranges between 0 and 1. |
Spatio-temporal cluster | In this article, a spatio-temporal cluster of
GPS points is defined as a group of GPS points which exhibit both spatial clustering and temporal continuity. |
Sub-location | During a staying activity, the subject may
actually stayed at several locations (for example, a subject’s staying at home activity may include sub-locations in the living-room and the yard) within the geographic range. Correspondingly, GPS points of the staying activity may exhibit several spatio-temporal clusters which denotes these sub- locations. |
The reason why we emphasize outdoor for walking activities is that when an individual walks inside a building (for example, walking inside a gym, shopping at Walmart), the GPS points may still exhibit a cluster around the building location, which makes it difficult for algorithms to distinguish. Therefore, indoor walking would be categorized into staying activities.