The Semantic Explicating of Spatial Data in Non-Relational Databases Case Study: Deteriorated Urban Environments

The World Wide Web Consortium (W3C) defines the semantic web as a framework to share and reuse of data between groups, communities and applications. In this relation Tim Berners-Lee proposed the idea of “web of data” that can be processed by machines. In order to reach the aims of semantic web, there are needs to data representation languages, knowledge bases and rule bases that are known as semantic web technologies. In the world of spatial data, the semantic information has important roles in different activities of spatial analysis and data management. In this research the spatial semantics are dealt with a difference in two levels. The first is all the semantics about the location and geometry of an individual object and the second is about the arrangement and mixture of objects in an extent. The latter is more related to spatial relations that cause complexities to cope with. The semantic explicating of such concepts in relational spatial databases has some limitations such as non-flexibility of database schema, non-flexibility of data types, the restricted defining of sub-classes, implicit semantics of relations and finite number of joins between tables. The exclusive characteristics of semantic web technologies in storing, managing and retrieving of data and knowledge lead to a wide range of applications; also the applications independent of web environment. The databases defined in semantic web technologies are based on graph structures and present the advantages of logical databases, object-oriented data bases and network databases. The introduced capabilities are applied to explicate the concept of "deteriorated structures" in urban environments. This concept is inherently non-spatial but has two complementary spatial interpretations; permeability and fine granularity; which make it possible that the concept to be explicit in spatial data of urban layouts. The scenario which based on, the semantic of " deteriorated area" is made explicit starts from a point, that by acquiring its coordinates it is determined whether this point is located within the such area or not? Handling of the spatial data of the case study in district 12 of Tehran and the procedure to semantic explicating of “deteriorated area” are done in the spatial and semantic extension of Oracle DBMS. The spatial data of urban parcels and streets accompanied with area of parcels and width of streets are used. In order to model the spatial relations, all the neighborhood relations between parcels and intersection relations between streets are also entered as linked data to triple store. In addition to RDF data, a rule base to exploit the concepts of “connected streets” and “co-blocked parcels” and also to define the conditions of narrow streets and small parcels is needed. Using some SPARQL queries the percentages of parcels with less than 200 m2 area and the streets with less than 6 m width are presented. The infinite number of inter tabular joints which have many applications in the spatial relations, is the main shortcoming of conventional databases and means we cannot express an infinite, recursively defined sequence of similar expressions. The procedure shows the deficiency of relational spatial database about recursive relations and implicit relations especially the spatial ones can be satisfied by semantic web data models and knowledge management techniques.