How I use Twitter for Research

It didn’t take too long to realize there are just too many papers to read, and it was frustrating to observe that the more you read the more is waiting to be read. So I (partially) solved this problem by following twitter accounts.

What I found especially useful on twitter is that (1) people actually still use it, and (2) among those who do, they’re really devoted to spreading the news they care about.

I follow #HCI and #LearningatScale related conferences (mainly #CHI, #UIST, #CSCW, #LearningatScale) and researchers, as well as departments and schools.

I also follow #MachineLearning related accounts too, but I tend to be very selective, because of the way the field has evolved… Mathieu Blondel and Ian Goodfellow are my two favorite ML related people/accounts. #ArXiv’s ML/Stats feed can be helpful, but it’s probably overwhelming if you want to read everything that pops up.

The general strategy is that I follow whatever account that looks interesting, and after a few days, if the account doesn’t provide anything personally interesting, I unfollow. And I have the luxury of being a nobody in the field that I don’t feel guilty when I decide not to follow someone anymore because it doesn’t offend them in any way. Ha!

Here's what I do every day, when I wake up I skim through my twitter feed and retweet everything (hence save it on my timeline) that I feel like I need to read at some point. Then later in the day I just need to go over my timeline and read/study the retweeted items one by one. I un-retweet if the content is not what I had thought or if it’s not interesting to me.

After having a pretty good list of following twitter accounts, by investing about 10 minutes every morning, I get to keep up with the latest research and other fun research results around the world.