AI question answering from Corona dataset released by Kaggle/from set of reliable News links

[lawlietlight]

Hi,
We have a question answering engine hosted in our dev box. It needs GPU and we would like to help the community by providing that api for free. It can attend to question answering using common English from a English data corpus almost always.
We can use it to narrow down the search space for people by searching relevant question by context and return the set of related or concrete questions with answers. Let us know If HertzAI can be of any help to implement this solution right away.

Link to dataset https://www.kaggle.com/sudalairajkumar/novel-corona-virus-2019-dataset/kernels

[bodhish]

@tomahawk-pilot @vigneshhari @gigincg @prometheansacrifice
What do you guys think?

[lawlietlight]

@tomahawk-pilot @vigneshhari @gigincg @prometheansacrifice
What do you guys think?

We can show a demo of our capability if you guys are interested.

[bodhish]

Would be great if you could share a demo.

You can connect to me on our public slack slack.coronasafe.in (@bodhi in slack)

[lawlietlight]

Guys we'll give them a demo for them to use our QA api.Regards,Sathish -------- Original message --------From: Bodhish Thomas <notifications@github.com> Date: 3/21/20 21:46 (GMT+05:30) To: coronasafe/awareness <awareness@noreply.github.com> Cc: lawlietlight <sathish@hertzai.com>, Author <author@noreply.github.com> Subject: Re: [coronasafe/awareness] AI question answering from Corona dataset released by Kaggle (#19) Would be great if you could share a demo. You can connect to me on our public slack slack.coronasafe.in (@bodhi in slack) —You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or unsubscribe. [ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "#19 (comment)", "url": "#19 (comment)", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } } ]

[dauntlessnomad]

how authentic is the data

[lawlietlight]

The question answering Engine is not data dependent, we can send any premise of text corpus from any news feeds, PR, articles which hosts lots of question answering and answer from the text corpus. We can manually cull all unprovable data and source question answers from reliable links like WHO/CDC/Indian central/state government websites alone.

The site below already culls the fake data or illegitimate numbers and verifies the data posted in their site as and when things happen.
https://www.worldometers.info/coronavirus/

I just checked https://www.mohfw.gov.in where we can answer state wise information.