Patient-Case-Similarity

Calculation of Patient Similarity based on Patient Demographic and Case Details extracted from XML annotations, Electronic Health Record (EHR)

Before running the code remember to delete the contents of the folder txtData

There may be more than one XML file for each patient, remember to concatenate them before running the code, the cat command will be of great use to you ...

• XSLT used for transforming and extracting annotated data into CSV.

• An ensemble model consisting of both Word Mover’s Distance (WMD) and General Feature Extraction

based on curated list of important sections were utilized in ratio 3:1.

Project for Grand Finale of Smart India Hackathon 2019 for the Problem statement by EzDI in which we made a system for assessment of similarity between patient using vector concatenation techniques and optimized WMD to foresee prognosis and hospitalization.