China
Africa
Explore chapters and articles related to this topic
Data driven life cycle management
Published in Airong Chen, Xin Ruan, Dan M. Frangopol, Life-Cycle Civil Engineering: Innovation, Theory and Practice, 2021
Linked Data is a web technology. Data from traditional databases are published in a web-format, the so-called “triples”. A triple consists always of 3 components: subject-predicate-object (RDF statement). A triple store or RDF store is a purpose-built database for the storage and retrieval of triples through semantic queries. Each object has a unique web address (a URI). Relations can come from the original data source, or can be defined to relate multiple data sources. Sometimes relations can be derived, for instance from the RWS-OTL or from the CB-NL. But With the help of AI-techniques, it is also possible to find relations that where not known at forehand (Artificial Intelligence, AI). Figure 3 shows the principle. Three data sets contain data of a same bridge, each in a different way using their own semantics. Once published as Linked Data, the datasets can be related. In this example, different sort of relations can be found between the sources: – Identical identifiers,– Semantic relations (“truss bridge”, “traffic bridge” and “steel bridge” are subtypes of “bridge”– Location– Common properties (The property “owner” can be mapped, but in this case the values are different. It might be “Piet” from the province “Zuid Holland”)
Semantic Technologies as Enabler
Published in Sarika Jain, Understanding Semantics-Based Decision Support, 2021
Relational databases are a rigid, structured, and well-established methodology for storing structured data, as compared to their NOSQL counterparts, which are document oriented. NOSQL graph databases store and manage data that is represented as an unlabeled or undirected or weighted graph. An RDF database is a specialized database for RDF triples, also called a triple store. Triple stores are a type of NOSQL graph store that store a special directed labeled graph, i.e., an RDF graph. Triple stores also use URIs required for querying and reasoning. They offer several advantages compared to traditional RDBMSs:
Real-Time Search in the Sensor Internet
Published in Ioanis Nikolaidis, Krzysztof Iniewski, Building Sensor Networks, 2017
As most Web content lacks semantic annotation, crawlers of semantic search engines, such as DBPedia,§ a semantic encyclopedia, concentrate on sources of structured data. Instead of keyword extraction and maintenance of an inverted index, gathered triples are usually stored in a specialized database, called triplestore. SPARQL, a query language for the graph formed by the RDF data, is used to retrieve information from the triplestore.
Knowledge on-demand: a function of the future spatial knowledge infrastructure
Published in Journal of Spatial Science, 2021
Lesley M. Arnold, David A. McMeekin, Ivana Ivánová, Kylie Armstrong
RDF Triple Stores utilise a Linked Data interface to configure access to the data store via SPARQL endpoints; and static data files, such as CSV files, excel spreadsheets and XML files, can be converted to RDF triples and served directly to the Web. Natural language text documents can be encoded as RDF using a Linked Data entity extractor. In addition to publishing data, the Publishing Layer includes a Knowledge Representation Repository to share domain ontologies, vocabularies, data dictionaries and metadata, all of which are used on the Web to represent knowledge – ranging from large taxonomies categorising Web sites to the classification of products for sale e.g. Amazon.com.