I'm increasingly focusing on the GIS side of misinformation analytics for the next phase of COVID https://t.co/0bUpNy4g7N -- if that sounds like your jam and you have volunteer time, plz DM or ping on our Slack :)
So much data ❤️💪 in this morning's https://t.co/0bUpNy4g7N analyst sync. We've started our first report, and in parallel, making it a tool for public health & civic leaders across the US to get their health messages out to at-risk folks.
If you do data, volunteers wanted!
The https://t.co/OPhLw5Wc9X data pipeline is live, and we've moved on to the data science side! First up are fighting clinical disinformation and large-scale digital crime.
The first data science & analysis Zoom is Sat @ 10a PT, volunteers welcome (see site.) See you there!
The https://t.co/OPhLw5Wc9X data pipeline is live, and we've moved on to the data science side! First up are fighting clinical disinformation and large-scale digital crime.
The first data science & analysis Zoom is Sat @ 10a PT, volunteers welcome (see site.) See you there!
Good news: https://t.co/r7OdY8BmuX COVID behavior change & anti-misinfo pipeline is now ingesting at 100+ events / s and we are starting to move on to GPU analytics for our first intervention.
❤️ To folks jumping into our COVID effort https://t.co/0bUpNy4g7N via @neo4j's #Graphs4Good hackathon!
We can use more 💪 on:
- Data eng/dev/sci: entity extractors, twitter alert bots, & misinfo scoring pipeline
- Intervention: fake drug misinfo
- Intervention: scams
- Logo 🌈
@KLdivergence We're ramping up on identifying localized social networks exhibiting risky & misinformed behavior and facilitating direct intervention through local leaders, akin to HIV interventions in 80's/90's.
Except we're computer people. Public policy people wanted :)
Signup on https://t.co/CVAFDxm3V5 to get the Trump cyborg detection session & code, and for when the COVID misinfo graph neural net session gets announced. (Contributors welcome!)
The story from our Ride Along on Trump "cyborg" bots continues w/ great work by Jess & Cody!
For more on data & detection, signup for Part II where we'll track COVID misinfo - a big problem now - w/ Graph Neural Nets.
https://t.co/34MVAcCBhx
Live in 1hr: The next Data Ride Along on "US Election Twitter bots & how to find them"!
Stream+chat: https://t.co/v03Qs5LkE1
Signup: Otherwise, receive post-show recordings+code, and announces of next shows: https://t.co/0W59oGu7TC
Live in 1hr: The next Data Ride Along on "US Election Twitter bots & how to find them"!
Stream+chat: https://t.co/v03Qs5LkE1
Signup: Otherwise, receive post-show recordings+code, and announces of next shows: https://t.co/0W59oGu7TC
"US Election Twitter bots and how to find them" is on for tomorrow!
We've been busy hunting election Twitter bots and preparing the DIY tools, so it'll be quite the show. Quick link for the live stream etc: https://t.co/0W59oGu7TC
Ride #3: US Election Twitter bots and how to find them using GPUs and graph APIs, with guests Cody Webb & Marc Singer (@nodexl@smrfoundation )
Join for the live interview and coding, sign up to get the recording + DIY notebook, and help by sharing!
https://t.co/NsfvTx5FMP
1. Monday is https://t.co/NdoCNT1Wrt on "Election twitter bots and how to find them!"
2. Any algorithm requests for the next one?
( Please share for more ideas! )
Ride #3: US Election Twitter bots and how to find them using GPUs and graph APIs, with guests Cody Webb & Marc Singer (@nodexl@smrfoundation )
Join for the live interview and coding, sign up to get the recording + DIY notebook, and help by sharing!
https://t.co/NsfvTx5FMP
@cjnolet@RazibObaid@leland_mcinnes@YALEBBS @NurTaz_Rahman @yalelibrary@DataRideAlongs 💯! We label umap clusters/neighbors by only keeping neighbor 'b' for sample 'a' if is "close" wrt the n_neighbor parameter ('k=15'): "neighbor(a,b): knn_dist[a,b] < c * knn_w[*,k].stddev()". Helps dirty data _a lot_!
If RAPIDS had df[*,k].mode()... :)