Fact checker based on wikipedia corpus, StandfortNLP, TagMe and WordNet.
Steps for running the application
- start docker container
- run spring
- 'access via UI' OR 'access the demos (training & test)'
- java 11
- docker
- docker-compose
- maven 3.1 or higher
Starts the docker containers necessary for the fact check.
docker-compose up
Maven command, for running the unit tests.
mvn clean test
Starts the spring boot application.
mvn spring-boot:run
Access the UI via browser (with enabled js).
- upload => GET http://localhost:8080/upload
- factCheck => GET http://localhost:8080/factCheck
Run the the fact checking with the training & test data files of the execise.
- factCheck training file (.ttl-output) => GET http://localhost:8080/demo/runFileTraining
- factCheck test file (.ttl-output) => GET http://localhost:8080/demo/runFileTest
- factCheck training file (.html-output) => GET http://localhost:8080/demo/runFactCheck
- ner training file (.html-output) => GET http://localhost:8080/demo/runNer
Internal API access.
- ping => GET http://localhost:8080/api/ping
- upload => POST http://localhost:8080/api/upload
- factCheck => GET http://localhost:8080/api/factCheck?sentence={text}
Internal WordNet API access.
- GET http://localhost:5679/synonyms/1/{word}
- GET http://localhost:5679/hypernyms/1/{word}
- GET http://localhost:5679/substance_meronyms/1/{word}
- GET http://localhost:5679/hyponyms/1/{word}
- GET http://localhost:5679/antonyms/1/{word}
- GET http://localhost:5679/substance_holonyms/1/{word}
- GET http://localhost:5679/causes/1/{word}
- AlkApiController: for the API calls, to running a "real" fact check
- AlkController: for delivering the UI
- DemoController: for running the test & training datasets and provide a download of the result