someshkar / covid19india-cluster

:microscope: Covid19 India Cluster Graph
https://cluster.covid19india.org
MIT License
989 stars 654 forks source link

"Admitted", "Indigo" etc. Showing Up As Travel Clusters Among Cities and Countries #147

Open ikartik90 opened 4 years ago

ikartik90 commented 4 years ago

Describe the bug Certain airlines such as "Indigo", "Emirates" and "Air Arabia G9", and words like "Admitted" and "Arrived" are showing up on the Travel Cluster as International Destination Nodes.

To Reproduce Steps to reproduce the behavior:

  1. Go to 'covid19india.org'
  2. Click on 'Clusters' tab
  3. Select 'Travel' cluster filter
  4. You would notice the above mentioned words showing up as international travel nodes on the graph.

Expected behavior These values should not show up on the map as they are either the means of travel or are actions, and have nothing to do with one's travel itinerary.

Screenshots

Screenshot 2020-03-28 at 2 48 44 PM Screenshot 2020-03-28 at 2 49 03 PM Screenshot 2020-03-28 at 2 49 30 PM

Additional context The problem is probably cropping up as this information is being fetched by crawling through the Notes tagged to each patient and exploring for comma separated values in those notes.

sibeshkar commented 4 years ago

This repo is using the NLP api for making structured travel data out of unstructured notes from https://github.com/NirantK/coronaIndia. Possible to report it on that repo?