NELL2RDF is a project to represent a valuable database in a standardized and self-describing dataset using Linked Data principles.
We are representing NELL(Never-Ending Language Learning) knowlegde base, a dataset with about 50 million candidate beliefs.
Our motivation is convert this valuable knowlegde base in a RDF. Becoming it more available and understandable to the community.

This paper was submitted to ISDWC 2017 Resource Tracks and can be accessed here.
We used Java and Jena for the implementation and all the project we developed during this aproach can be accessed in the github: https://github.com/WDAqua/nell2rdf

The datasets obtained as results are available as follows:

Beliefs Ontology: https://w3id.org/nell2rdf/ontology/nellrdf.ontology.n3
Provenance Ontology: https://w3id.org/nell2rdf/provenance/ontology/nellrdf.provenance.ontology.n3
Promoted Belief: https://w3id.org/nell2rdf/nellrdf.promoted.n3.gz
Candidate Beliefs: https://w3id.org/nell2rdf/nellrdf.candidates.n3.gz

Our work is supported by funding from the EU H2020 research and innovation program under the Marie Skodowska-Curie grant No 642795.
We would like to thank Bryan Kisiel from NELL's CMU team for his technical support about NELL's components.

see archive

José Gimenez Garcia
Maísa Cristina Duarte
Antoine Zimmermann
Christophe Gravier
Estevam R. Hruschka