patient-case-similarity is a Natural Language Processing (NLP) project used to calculate the similarity between two patients using WMD (Word Movers Distance).
The project was done as a part of Smart India Hackathon 2019 at National Institute of Technology Warangal, India.
Project depicted was presented at the national level of SIH 2019.
The user needs to provide the permission in linux Terminal using
chmod +x script.sh
After the permission is granted, user has to run the script.sh
If you're having trouble running the script, you can type the commands manually in Linux terminal present in script.sh file.
The sample data (given to us) is present ./'Sample Dataset'/NER_XML/
The file used in shell script is copied from here to the root directory.
xslt is used for transforming and extracting annotated data into .csv from .xml
The data from .xml file is extracted and stored in ./data/ directory. The extracted data is preprocessed using the pre_process.py and finally the WMD Score is calculated using wmd.py.