Even though there is plenty of public data, I cannot find a source for the global infection chains of Corona. In other words, I am interested in the data that shows from where each infection has been imported (e.g. country A's patient 0 has imported Corona from country B).

You can find information about that here and there in various news articles. However, I hope for a data source that can be used automatically.

2 Answers 2


I'm not sure anybody in the world has this information.

I live in Italy, not far from the largest coronavirus spreading in Europe. Patient 1 (the guy who started the epidemic in all Italy) used to live about one hour from my town. No one has still a clue on how he got it, and how it all started. I'm quite sure there's no such data at all. I'm sorry about this.

I suggest you to take the coronavirus dataset available on Kaggle and maybe try to estimate it by combining it with geo-data. That's the only thing I can think ok. Hope this helps.

  • Thanks for the link! Of course there is no such data. Otherwise the official institutions wouldn't struggle at all. Mapping confirmed cases and geo data won't work - I think - because all you know is when it pops up in some country. However, news say something like "first patient was travelling from xy country". So there is a strong possibility that it got imported from xy country.
    – AmonNascaroth
    Commented Mar 6, 2020 at 10:29
  • I see. In that case you might try to do it by scraping news websites, I guess? It's hard work, but the results would be extremely interesting.
    – Leevo
    Commented Mar 6, 2020 at 10:43

So on Kaggle there is a "line list" data source which is quite interesting. https://www.kaggle.com/sudalairajkumar/novel-corona-virus-2019-dataset

In this list you will find a column showing whether it's the first case in a particular country, if the patient was visiting or coming from Wuhan.

However, there are cases where it's not that binary. Take Germany for example:

new confirmed COVID-19 patient in Germany: 1/28-No.1 male, 33, caught from Chinese colleague during conference in Munich from 1/20-1/21, first human-to-human transmission in Europe, confirmed 1/27/202...

Or Italy

new confirmed COVID-19 patient in Rome, Italy: male, Chinese tourist, arrived in Milan over a week ago

Both are first cases, so therefore the number of cases is 1 and can be filtered easily. However, there are nth confirmed cases which have been imported too. Therefore, some automation filtering those entries would be helpful.

Could NLP help? I am a novice in these topics and started learning, so excuse me if I haven't tried yet.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.