search

dice-group / AGDISTIS

AGDISTIS - Agnostic Named Entity Disambiguation
http://aksw.org/Projects/AGDISTIS.html
GNU Affero General Public License v3.0
141 stars 37 forks source link

Minimal Ontology Example not working #46

Open RicardoUsbeck opened 7 years ago

RicardoUsbeck commented 7 years ago

Given the knowledge base below:

http://fairhair.ai/kg/resource/Evertec http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://dbpedia.org/ontology/Company .
http://fairhair.ai/kg/resource/Evertec http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://dbpedia.org/ontology/Organisation .
http://fairhair.ai/kg/resource/Evertec http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://schema.org/Organization .
http://fairhair.ai/kg/resource/Evertec http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://www.ontologydesignpatterns.org/ont/dul/DUL.owl#SocialPerson .
http://fairhair.ai/kg/resource/Evertec http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://www.wikidata.org/entity/Q43229 .
http://fairhair.ai/kg/resource/Evertec http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://dbpedia.org/ontology/Agent .
http://fairhair.ai/kg/resource/Evertec http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://www.ontologydesignpatterns.org/ont/dul/DUL.owl#Agent .
http://fairhair.ai/kg/resource/Evertec http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://www.w3.org/2002/07/owl#Thing .
http://fairhair.ai/kg/resource/Evertec http://www.w3.org/2000/01/rdf-schema#label "Evertec"@en .
http://fairhair.ai/kg/resource/Evertec http://dbpedia.org/ontology/address "Cupey Center Building Road 176"@en .
http://fairhair.ai/kg/resource/Evertec http://dbpedia.org/ontology/numberOfEmployees "1660"^^http://www.w3.org/2001/XMLSchema#nonNegativeInteger .
http://fairhair.ai/kg/resource/Evertec http://dbpedia.org/ontology/locationCity http://dbpedia.org/resource/San_Juan,_Puerto_Rico .
http://fairhair.ai/kg/resource/Evertec http://dbpedia.org/ontology/location http://dbpedia.org/resource/Puerto_Rico .
http://fairhair.ai/kg/resource/Evertec http://dbpedia.org/ontology/locationCountry http://dbpedia.org/resource/United_States .
http://fairhair.ai/kg/resource/Evertec http://dbpedia.org/ontology/industry http://dbpedia.org/resource/Information_technology .
http://fairhair.ai/kg/resource/Evertec http://dbpedia.org/ontology/foundingYear "2004"^^http://www.w3.org/2001/XMLSchema#gYear .

AGDISTIS should be to disambiguate the following sentence 'Evertec is a company in Puerto Rico.'

Currently it returns the "don't know" links for both entities.

Please write a unit test for it.

DiegoMoussallem commented 7 years ago

It worked from my side, however, we have to make AGDISTIS handle more than node type in the KB. I have alleviated this problem by commenting the line 225 in CandidateUtil.java, but I'm absolutely sure that it will not work for complex graphs when performing HITS or PageRank.

RicardoUsbeck commented 7 years ago

Unit test or it did not happen :) And commenting is not a good solution since there might be other cases where we need that line. So we should keep that issue open

DiegoMoussallem commented 7 years ago

I think I have mixed two problems here. I'm gonna open another issue to handle more than one nodeType. However, I don't see it as a bug. Download it and run pls https://drive.google.com/open?id=0BwQZNbPH4KyVRFRiZDVEZmpSZHM

RicardoUsbeck commented 7 years ago

Thanks for the link. Can we have a unit test loading this data, indexing it and then running the query?

RicardoUsbeck commented 6 years ago

So I transformed it working TTL using http://ttl.summerofcode.be/:

<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://dbpedia.org/ontology/Company>.
<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://dbpedia.org/ontology/Organisation>.
<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://schema.org/Organization>.
<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.ontologydesignpatterns.org/ont/dul/DUL.owl#SocialPerson>.
<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.wikidata.org/entity/Q43229>.
<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://dbpedia.org/ontology/Agent>.
<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.ontologydesignpatterns.org/ont/dul/DUL.owl#Agent>.
<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.w3.org/2002/07/owl#Thing>.
<http://fairhair.ai/kg/resource/Evertec> <http://www.w3.org/2000/01/rdf-schema#label> "Evertec"@en.
<http://dbpedia.org/resource/Puerto_Rico> <http://www.w3.org/2000/01/rdf-schema#label> "Puerto Rico"@en.
<http://fairhair.ai/kg/resource/Evertec> <http://dbpedia.org/ontology/address> "Cupey Center Building Road 176"@en.
<http://fairhair.ai/kg/resource/Evertec> <http://dbpedia.org/ontology/numberOfEmployees> "1660"^^<http://www.w3.org/2001/XMLSchema#nonNegativeInteger>.
<http://fairhair.ai/kg/resource/Evertec> <http://dbpedia.org/ontology/locationCity> <http://dbpedia.org/resource/San_Juan,_Puerto_Rico>.
<http://fairhair.ai/kg/resource/Evertec> <http://dbpedia.org/ontology/location> <http://dbpedia.org/resource/Puerto_Rico>.
<http://fairhair.ai/kg/resource/Evertec> <http://dbpedia.org/ontology/locationCountry> <http://dbpedia.org/resource/United_States>.
<http://fairhair.ai/kg/resource/Evertec> <http://dbpedia.org/ontology/industry> <http://dbpedia.org/resource/Information_technology>.
<http://fairhair.ai/kg/resource/Evertec> <http://dbpedia.org/ontology/foundingYear> "2004"^^<http://www.w3.org/2001/XMLSchema#gYear>.

Note: I added a label for Puerto Rico