Geospatial Semantic Query Engine for Urban Spatial Data Infrastructure

Introduction

The rapid technological evolution is characterizing diverse disciplines like disaster management, urban planning and smart cities is becoming a key factor to trigger true user driven applications. The 3D city models are digital models of urban areas that represent terrain surfaces, buildings, vegetation, etc. (“3D city models,” 2016). In this context, 3D city models will play an increasingly important role in modern city spatial information infrastructure. The basic steps that are followed for various applications are: encode, store and manage. There are several 3D city model encoding methods are available in the literature that can be used to create and save the 3D city models in GeoTIFF, AutoCAD, .3DS, .OBJ, Collada and KML etc. (“3D file format,” 2012). The 3D city models are stored and managed using both file and database systems. Database approach is a good method for continuous management and distribution of 3D city models (Mao et al., 2014). Database systems facilitate scalability, stable and reliable management for both spatial and non-spatial data. Spatial database systems that support 3D city geometry can be used for storing and managing 3D city models. However, due to heterogeneity and diversity in the 3D city models, there is no single and unique schema to store and manage 3D city data.

To address the heterogeneity and diversity in the 3D city models, OGC introduced a standard called CityGML (“CityGML,” n.d.). CityGML is an international standard defined by Open Geospatial Consortium (OGC) for the representation and exchange of 3D city models (Gerhard & Lutz, 2012). CityGML allows users to employ virtual 3D city models for advanced analysis and visualization tasks in diverse urban geospatial applications. CityGML plays a leading role in the modularization of urban geospatial information. The 3D City Database (3DCityDB) (“3DCityDB,” 2016) is a 3D geo database that stores and manages virtual 3D city models. The database model contains semantically rich, hierarchically structured, multi-scale urban objects facilitating complex GIS modeling and analysis tasks. 3DCityDB-Web-Map-Client is a web-based front-end 3D visualization and interactive exploration of large semantic 3D city models. However, querying multiple 3D city objects requires multiple joins that increases query complexity and impact application performance, exploring spatial topological and semantic relationship between all or few city objects requires multiple queries and users who are not familiar with database technologies and 3DCityDB relational schema cannot do spatial analysis on 3D city objects.

There is an important need to develop an information infrastructure product which is GENERIC, SIMPLE, friendly to novice user. This facilitates automatic establishment of spatial and semantic relationship between all city objects by a simple, generic, spatial and semantic query engine on 3D city data with ontology support. This gives an opportunity for a strong dynamic GeoSPARQL query building and execution algorithms considering all the possible ontology element driven query patterns. In general, the identification of frequent sample query patterns of 3D city data users utilizes the typical areas of applications such as disaster management, urban planning and smart city development explained conceptually as follows: Disaster management typical queries would be “visualize and show all school buildings with height greater than 30 feet within flooded area” needing urgent attention and evacuation. Urban planning typical queries like “find out all the city objects nearby a State or National Highway. Smart city typical queries would be “find out streetlamps within commercial land and residential area”. Additionally, users can find all the city objects “nearby” a power tower for 3D visualization. This forms the motivational driving force in this design and development of the current research work.

Applications of CityGML need effective spatial semantic analytics and query system on CityGML data. CityGML application performance is improved through ontology driven spatial semantic query engine. Ontology driven querying system enhances the querying facility on:

• •

  Every city object (Giving city object details and its spatial and spatial semantic relations with other city objects);

• •

  City object class of a CityGML module (Abstract class e.g. Transportation);

• •

  Multiple City object classes (E.g. Road) of a CityGML module;

• •

  Multiple classes of multiple CityGML modules;

• •

  City object class with single attribute;

• •

  City object class with multiple attributes;

• •

  City object with spatial relations (e.g. within, contains, etc.);

• •

  City object with spatial semantic relations (e.g. nearby, neighbor, etc.).