Querying probabilistic land cover data using fuzzy set theory

Queries expressed in verbal language often involve a mixture of uncertainties in the outcomes of events that are governed by chance and in the meaning of linguistic terms. This study exemplifies how probability and fuzzy sets can work together to deal with such queries in the spatial domain. It involves site selection on the basis of accessibility (travel time) estimates and per-pixel probabilities of land cover change derived from remotely sensed imagery. Relationships between probabilities and fuzzy sets were established on the basis of a linguistic probability qualifier (high probability) and the expectation of a membership function defined on stochastic travel time. Fuzzy query processing was compared with crisp processing to emphasise the difference between grade and probability of membership. Fuzzy set theory is used to deal with the vague meanings of linguistic terms. The fuzzy query response contained more information than the crisp response. namely the degree to which individual locations matched the selection criteria. This illustrates the gain in expressive power provided by combining probability and fuzzy sets.