search

LinuxForHealth / patient-browser

Other
3 stars 7 forks source link

Issue 10: Additional context around the text span for NLP #27

Closed mcbride-p closed 2 years ago

mcbride-p commented 2 years ago

Code changes to show additional context (up to 30 characters) around the text span for NLP and highlight the actual span

mcbride-p commented 2 years ago

@ewill00 - To answer your questions above ... Yes const value determines character extension on either side and agree this should be configurable and determined from a config file that can be changed on deployment going forward. I'll create ticket for that. It does handle scenarios where there are in excess of 30 characters before and after the startoffset and endoffset respectively. In either case span text is preceded/superceded by three dots after the 30 characters to indicate evidence_detail text extends further. Correct - No we dont handle ending extended text on word boundary.