This project has been developed within the Open Science and Artificial Intelligence in Research Software Engineering class.
This framework shows the power of the pdf parser grobid in combination with different xml parser by showing result for the following questions for scientific papers provided by the user.

1. Keywordcloud of the abstract
2. Number of figures
3. List of links

##How to start

1. Make sure you have GROBID installed and running

• For example by using docker. Run:
  docker pull lfoppiano/grobid:0.7.2
  to install and
  docker run -t --rm -p 8070:8070 lfoppiano/grobid:0.7.2
  to start GROBID
• For further information see the documentation

2. Set up your environment

• By running conda create --name <env> --file requirements.txt

3. Provide your pdfs and run

• Move your pdf to the resources directory OR choose your own directory by changing the RESOURCE_DIRECTORY constant in xml_parser.py
• Start the application by running python xml_parser.py

If you want parse existing xml files. Set the MODE constant in xml_parser.py to "XML" provide your xml files in a directory and assign your directory to the XML_RESSOURCE_DIRECTORY constant in xml_parser.py

• Now just start the feature extraction for xml files by running python xml_parser.py

See full documentation at https://os-ai-cd.readthedocs.io/en/latest/