• Python 2.7
• ElasticSearch
• apache-jena-fuseki-3.4.0

• Download or clone the repository $ git clone https://github.com/jtrillos/QAUniBonn.git
• Download and install ElasticSearch
  • Start ElasticSearch $ sudo service elasticsearch start
• Download and extract the apache-jena-fuseki-3.4.0
  • Go to the config.ttl in the QAUniBonn repository and modify the path in line 21
  • Run Fuseki $ ./fuseki-server -config=/path/to/the/repository/QAUniBonn/Docker/Fuseki_TripleStore/configuration/config.ttl
  • The server starts at http://localhost:3030/
• Create a virtual environment for Python 2.7 $ virtualenv pythenv --python=python2
  • Run the virtual environment $ source pythenv/bin/activate
  • Install the dependencies $ pip install elasticsearch SPARQLWrapper simplejson gensim pandas flask nltk $ pip install -U scikit-learn
• Run $ python IndexSDAKG.py to import all the entities and labels of the KG to ElasticSearch (This is needed only when the KG is being updated or it is the first time).
• Download wiki-en.bin fasttext
  • Move the wiki.en.bin file into the data folder of the project.
• If there is not data of nltk uncomment the lines 8 - 11 in nerQuestion.py

To run the server

$ python app.py

To run the server in background

$ nohup python app.py &

$ curl -i -H "Content-Type: application/json" -X POST -d '{"question":"Where is SDA?"}' http://localhost:8000/ask