The Named Entity Tagging Engine: linking text annotations to (external) datasets of entities
The Entity Linking Engine uses Referenced Sites to search for Entities based on given Text Annotations.
Configuration
The configuration decides, which dataset you want to use as linking target. The default value is "local" referencing to the default DBpedia index. You may also decide on whether given types should restrict the set of possible links. E.g. for DBpedia, some organisations are not tagged as such, therefore, you want get them with this engine although, you expect them from your dataset.
- Referenced Site: {local, your referenced site}: The ID of the Entityhub Referenced Site used for semantic lifting of TextAnnotations.
- Persons: {true, false}: Set to TRUE to enable semantic lifting of Persons
- Person Type {
, dbp-ont:Person}: The rdf:type used to search for Persons. If empty Entities of any type are accepted. - Organisations {true, false}: Set to TRUE to enable semantic lifting of Organisations
- Organisation Type {
, dbp-ont:Organisation}: The rdf:type used to search for Organizations. If empty Entities of any type are accepted. - Places {true, false}: Set to TRUE to enable semantic lifting of Places
- Place Type {
, dbp-ont:Place}: The rdf:type used to search for Places. If empty Entities of any type are accepted.* - Label Field {
, rdfs:label}: The field used to search for Entities with a label similar to the selected text of the Text Annotation. If empty rdfs:label is used as default*
Example Result
For the sentence "The Stanbol enhancer can detect famous cities such as Paris and people such as Bob Marley.", you will get several EntityAnnotations for the terms "Paris" and "Bob Marley" from your linking target resource (in this case DBpedia) together with a confidence value, which can be used to sort the suggestions, e.g.:
{
"@subject": "urn:enhancement-b98283ae-845d-6666-d68b-f649852a7959",
"@type": ["enhancer:Enhancement","enhancer:EntityAnnotation"],
"dc:created": "2012-02-29T11:34:56.383Z",
"dc:creator": "org.apache.stanbol.enhancer.engines.entitytagging.impl.NamedEntityTaggingEngine",
"dc:relation": "urn:enhancement-b3d4617d-1760-0374-f471-e0e746003f4e",
"enhancer:confidence": 16641.191,
"enhancer:entity-label":
{
"@literal": "Bob Marley",
"@language": "en"
},
"enhancer:entity-reference": "http://dbpedia.org/resource/Bob_Marley",
"enhancer:entity-type":
["dbp-ont:MusicalArtist", "foaf:Person", "dbp-ont:Artist",
"dbp-ont:Person", "owl:Thing"],
"enhancer:extracted-from": "urn:content-item-sha1-37c8a8244041cf6113d4ee04b3a04d0a014f6e10"
},
or
{
"@subject": "urn:enhancement-785a4c4f-dc7d-aa46-91a2-aef840542ae2",
"@type": ["enhancer:Enhancement","enhancer:EntityAnnotation"],
"dc:created": "2012-02-29T11:34:56.383Z",
"dc:creator": "org.apache.stanbol.enhancer.engines.entitytagging.impl.NamedEntityTaggingEngine",
"dc:relation": "urn:enhancement-c176f1bf-a1dd-830e-df7d-deecdfdc8375",
"enhancer:confidence": 1323049.5,
"enhancer:entity-label":
{
"@literal": "Paris",
"@language": "en"
},
"enhancer:entity-reference": "http://dbpedia.org/resource/Paris",
"enhancer:entity-type":
["dbp-ont:PopulatedPlace","dbp-ont:Settlement",
"http://www.opengis.net/gml/_Feature",
"dbp-ont:Place","owl:Thing"],
"enhancer:extracted-from": "urn:content-item-sha1-37c8a8244041cf6113d4ee04b3a04d0a014f6e10"
}