Ontology-Driven Translation of Geospatial Data

 

This web site provides the implementations accompanying my PhD Thesis. These consist of:

 

Chapter 4 - Semantic Reference Frame for Geospatial Data Models

 

All provided ontologies have been developed in WSML-Flight. They can (for example) be displayed as graph, text, and in a proprietary ontology editor view using the Web Service Modelling Toolkit (WSMT).

 

ZIP-file with all

 

DOLCE Classes

(A profile of DOLCE for establishing a semantic reference frame for geographic data.)

 

DOLCE Mereology

(A profile of DOLCE for establishing a semantic reference frame for geographic data, mereological relations between Endurants, we do not focus on time, for this reason we simplify the part-hood of DOCLE from P(x,y,t), where x and y are endurants and t is a time interval to P(x,y).)

 

DOLCE Lite Plus (DLP)

(This ontology captures foundational knowledge, it implements parts of the formal DOLCE extension DOLCE Lite Plus focussing on translation requirements for semantic reference frames.)

 

DOLCE Spatial Features

(This ontology captures the domain knowledge for (spatial features), it implements parts of the formal DOLCE extension from PROBST focussing on translation requirements for semantic reference frames, especially parts required for disambiguitising data models for road width, this can of course be generalised.)

 

DOLCE Qualities and Quality Spaces

(This ontology captures the domain knowledge for qualities and quality spaces, it implements parts of the formal DOLCE extension from PROBST focussing on translation requirements for semantic reference frames, especially parts required for disambiguitising data models for road width, this can of course be generalised.)

 

DOLCE Reference Spaces

(This ontology captures the domain knowledge for reference spaces, it implements parts of the formal DOLCE extension from PROBST focussing on translation requirements for semantic reference frames, especially parts required for disambiguitising data models for road width, this can of course be generalised.)

 

DOLCE Geospatial Features

(A generalisation of the categorisation of boundaries to the notion of DOLCE features.)

 

DOLCE Region Connectivity Calculus (RCC)

(An ontology about RCC for DOLCE features and extensions. Including derivation rules from DOLCE mereology.)

 

Semantic Datums Instances

(Most commonly used semantic datums, current version: metric datums only.)

 

Geospatial Ontology

(This ontology captures the domain knowledge for the gespatial domain, it formalisis parts of the DOLCE extension from KLIEN focussing on translation requirements for semantic reference frames, especially parts required for disambiguitising data models from the road administration domain, this can of course be generalised.)

 

Annotation

(This ontology captures the old and new relations for implementing annotations.)

 

Attribute Mappings

(This ontology captures the various allowed attribute mappings (lower bound, upper bound, and exact match.)

 

Augmentation Ontology

(This ontology captures inverse relation to kinds of translation operators. Those are used for backtracking applications when generating translation scripts.)

 

Haskell Translation Ontology

(This ontology captures the translation/augmentation operations, which are available from the specific Haskell implementation.)

 

XLink Ontology

(An ontology of XLink elements.)

 

XML Schema Ontology

(An ontology of XML Schema elements, due to project requirements we restrict this ontology to build-in data types.)

 

GML Ontology

(An ontology of the GML 3.1.2 elements.)

 

Road Ontology

(This ontology captures the most important notions when considering roads as part of a semantic reference frame for translation.)

 

ATKIS Feature Type Ontology (FTO)

(This ontology captures the structure of the sub-model of ATKIS and implements the proposed annotation.)

 

INSPIRE Feature Type Ontology (FTO)

(This ontology captures the structure of the sub-model of INSPIRE and implements the proposed annotation.)

 

Proof of Concept file

(This file imports all relevant ontologies for testing the developed reference frame for geospatial information on the ATKIS and INSPIRE FTOs.)

Chapter 5 - Algebraic Theory for Translation

 

The algebraic theory for translation has been developed with Haskell, using the Glasgow Haskell Compiler (GHC). Separate modules can be executed using the following command:

 

ghci -XMultiParamTypeClasses -XFunctionalDependencies -XFlexibleInstances -XUndecidableInstances

 

ZIP-file with all

 

Module: Observations or Measurement

(This module provides an extended value model for encoding measurement and observation results.)

 

Module: GML

(Module providing central GML (version 3.2.1) data types, which are extended with the new value model (ObservationOrMeasurement).)

 

Module: ATKIS Model

(Haskell representation of the ATKIS sub-model.)

 

Module: ATKIS Data

(Example Haskell data set for the ATKIS sub-model.)

 

Module: INSPIRE Transport Network Model

(Haskell representation of the INSPIRE Transport Networks (TN) sub-model as a target data model.)

 

Module: Overlay

(Module providing an implementation of interval overlay with propagation of imprecision.)

 

Module: Attribute Datums

(Defining a theory for attribute datums, overlap and translation is defined separately.)

 

Module: Augmentation

(Implementing a length calculation as augmentation operation. This includes means for error propagation and derivation of attribute datum.)

 

Module: ATKIS to INSPIRE

(Translation script from ATKIS sub-model to the INSPIRE TN sub-model.)

Chapter 6 - Proof of Concept

 

XML Schema with SA-WSDL annotations.

 

ATKIS Sub-Schema

(The GML conformant sub-model of ‘ATKIS Strassen’, extended with SA-WSDL.)

 

INSPIRE Transport Network Sub-Schema

(The GML conformant sub-model of ‘INSPIRE Data Specifications for Transport Networks’, extended with SA-WSDL.)

 

Demonstrator

(The demonstrator has been implemented in Java (Version 1.6). All required libraries and structures are included in the downloadable RAR file. It includes an executable JAR file and four folders. The ‘ontologies’ and ‘output’ folders are required for running the tool. The former hosts the required WSML ontologies, while the later is used for writing results. Here, we store the final translated data set in GML together with the dynamically created translation script. The folders ‘xsd’ and ‘data’ are included for easy access to required inputs. The used schemas are included in the ‘xsd’ folder, while the source data set is contained in the ‘data’ folder. These resources can be loaded from any other location, too.)

 

Example ATKIS data

(A GML conformant data set that applies the sub-model of ‘ATKIS Strassen’.)

 

Example ATKIS data in INSPIRE model

(The result of translating the example data set using the proposed approach.)