Formal Representation of Fuzzy Data Model Using Description Logic

Abstract

Geospatial information is becoming an integral part of many decision making processes, like, natural resource management, socio-economic development/planning, disaster management etc. However, the geospatial datasets are usually collected and managed by different organizations in their proprietary formats (or data models). Lack of interoperability between the datasets has become a major bottleneck for sharing and utilization of these heterogeneous spatial repositories. Thus there is a need for standardization of the geospatial data models (metadata) to facilitate interoperability among the heterogeneous repositories. The leading organizations use object oriented concept as standard for modeling spatial data. Further, the fuzziness is an intrinsic property of geospatial object. The existing metadata standards are meant for crisp spatial objects and fail to address the fuzzy properties. In order to describe geospatial objects more precisely, the fuzziness in these spatial objects should be captured and represented in data model (metadata). In general, UML is used as standard for data modeling using object oriented concept. However, expressiveness of the UML constructs do not have precise semantics, and are machine incomprehensible, and automated reasoning with UML is difficult. In this work, an attempt has been made to formalize the fuzzy geospatial data model, using description logic, to develop a fuzzy knowledge base, which may facilitate automated reasoning and sharing of spatial data across diverse repositories. The proposed work has been demonstrated by a running case study.