TY - JOUR
T1 - Positioning localities based on spatial assertions
AU - Liu, Y.
AU - Guo, Q. H.
AU - Wieczorek, J.
AU - Goodchild, M. F.
N1 - Funding Information:
We thank the reviewers for their comments that helped strengthen the paper. This research is supported by the BioGeomancer Project funded by the Gordon and Betty Moore Foundation, the National High Technology Development 863 Program of China (Grant No. 2007AA12Z216), and NSFC (Grant No. 40701134 and 40629001).
PY - 2009/11
Y1 - 2009/11
N2 - In practice, descriptive localities are often communicated using named places and spatial relationships. Uncertainty associated with such descriptions of localities is inevitable, and knowledge of such uncertainty is normally not explicit. When translating descriptive localities into spatially explicit ones, it is critical to circumscribe locations and to estimate the associated uncertainty based on a set of appropriate spatial relationships. In conventional research on qualitative spatial reasoning (QSR), spatial relationships are modeled using formal logic. Unfortunately, QSR cannot deal with the uncertainty of a position. In this paper, based on the conceptual model of spatial assertions, we introduce the uncertainty field model to represent the probability distribution of a point locality. Using probability operations, we can combine a set of assertions to position a locality. Conflicts among assertions for a single locality can be detected based on the resulting field. Since spatial relationships play an important role in the uncertainty of target objects, we investigate conceptually the uncertainty fields associated with various types of spatial relationships (for example, topological, directional and metric). In a concrete application, these uncertainty fields can be customized and used without altering the proposed framework.
AB - In practice, descriptive localities are often communicated using named places and spatial relationships. Uncertainty associated with such descriptions of localities is inevitable, and knowledge of such uncertainty is normally not explicit. When translating descriptive localities into spatially explicit ones, it is critical to circumscribe locations and to estimate the associated uncertainty based on a set of appropriate spatial relationships. In conventional research on qualitative spatial reasoning (QSR), spatial relationships are modeled using formal logic. Unfortunately, QSR cannot deal with the uncertainty of a position. In this paper, based on the conceptual model of spatial assertions, we introduce the uncertainty field model to represent the probability distribution of a point locality. Using probability operations, we can combine a set of assertions to position a locality. Conflicts among assertions for a single locality can be detected based on the resulting field. Since spatial relationships play an important role in the uncertainty of target objects, we investigate conceptually the uncertainty fields associated with various types of spatial relationships (for example, topological, directional and metric). In a concrete application, these uncertainty fields can be customized and used without altering the proposed framework.
KW - Geographic information system
KW - Probability
KW - Spatial positioning
KW - Spatial relationship
KW - Uncertainty field
UR - http://www.scopus.com/inward/record.url?scp=70350757453&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70350757453&partnerID=8YFLogxK
U2 - 10.1080/13658810802247114
DO - 10.1080/13658810802247114
M3 - Article
AN - SCOPUS:70350757453
SN - 1365-8816
VL - 23
SP - 1471
EP - 1501
JO - International Journal of Geographical Information Science
JF - International Journal of Geographical Information Science
IS - 11
ER -