A NLP proof-of-concept for doing automatic relationship extraction from texts.
APACHE-2.0 License
This scala program runs CoreNLP to extract basic relationships defined in a json such as "father", "mother" etc from a text. It was part of a university project course. See the article for more information.
The purpose of this program is to extract relationships between persons (named entities) in a text.
Detection of named entities is automatic and the relationships are defined by keywords and wikidata IDs in data/relations.json
.
Currently we support english with the intention of extending that support to at least another language, most likely Swedish.
The Relationship Extractor uses SBT for dependencies management and as a build tool. By default it uses the CoreNLP server to speed up multiple calls.
sbt compile
./run_server.sh
sbt run