Filtern
Erscheinungsjahr
- 2016 (4)
Dokumenttyp
- Beitrag zu einem Tagungsband (2)
- Vortrag (2)
Sprache
- Englisch (4) (entfernen)
Referierte Publikation
- nein (4) (entfernen)
Schlagworte
- Ontology (4) (entfernen)
Eingeladener Vortrag
- nein (2)
The amount of new knowledge about bridges, materials they are made of, deterioration processes for the different materials, resulting defects, methods for early detection of them and related causes for damage increases day by day. Often the knowledge remains in the heads of ageing specialists leaving the inspection teams or universities. This paper presents a feasibility study about application of a new tool from artificial intelligence to model and organize bridge engineering knowledge from a domain expert´s view. Exemplarily, knowledge and not only data from bridge testing in the field and in the laboratory were formalized for the presentation of the functionality of ontologies.
Bridges represent a crucial part of traffic infrastructure networks for both, freight transport and transport of people. None the less, sometimes it is not easy for infrastructure owners to manage or extract data and knowledge of interest about specific bridges from bridge management systems, e.g. along a line. In case of search for a special material parameter or for a recent strengthening measure, the selection of the wanted information may be time consuming. The search process depends on the complexity of the selected issue.
The paper presents the applicability of ontologies, a new tool from artificial intelligence transferred to bridge engineering issues. Current modern ontologies consider the rules from the World Wide Web consortium (W3C), use a defined Web Ontology Language (OWL) and the Dublin Core Metadata Initiative (DCMI®) to describe the members of a domain and the relations between them in a machine readable manor. Any information, data as well as text information is broken down to a simple triple (RDF) in a subject- predicate- object relation to organize the link between them. Advantage over any type of data base is processibility due to the use of restrictions and axioms. Visualization plug-ins, as OntoGraph or GraphViz, can visualize the structure of the ontology or parts of it and the described relations.
In the recent years, ontology editors like e.g. Protégé, were developed to enable domain experts to structure a specific domain. The editor is still under development. None the less, the study shows exemplarily, how an ontology for riveted steel bridges could consider results from testing in the laboratory and in the field. Besides typical materials, typical damage processes defects and structures, applicable methods are formalized, classified and described. Queries using e.g. DL-Query may address tasks of interest and extract it from the ontology.
The amount of new knowledge about bridges, about the materials they are made of, deterioration processes for the different materials, typical defects, methods for early detection of them and related causes for damage increases day by day. Often the knowledge remains in the heads of ageing specialists leaving the inspection teams or universities. This paper presents a feasibility study about application of a new tool from artificial in-telligence, ontology, for modeling and organizing bridge engineering knowledge from a domain expert´s view. Exemplarily, knowledge and not only data from bridge testing in the field and in the laboratory were formalized for the presentation of the functionality of ontologies. The open source ontology editor Protégé was ap-plied to test the feasibility.
The amount of new knowledge about bridges, about the materials they are made of, deterioration processes for the different materials, typical defects, methods for early detection of them and related causes for damage increases day by day. Often the knowledge remains in the heads of ageing specialists leaving the inspection teams or universities. This paper presents a feasibility study about application of a new tool from artificial in-telligence, ontology, for modeling and organizing bridge engineering knowledge from a domain expert´s view. Exemplarily, knowledge and not only data from bridge testing in the field and in the laboratory were formalized for the presentation of the functionality of ontologies. The open source ontology editor Protégé was ap-plied to test the feasibility.