Facebook Data Team: What’s on your mind? | Hacker News Reader

Facebook Data Team: What’s on your mind?

facebook.com

54 pointsisomorph15 comments

I love this OkCupid-esqe postings although I think Facebook could do a lot better but maybe they're concern about people's reactions to what they know.

Interesting point- I was about to comment that I loved this recent trend in intriguing data posts by Facebook. While I do understand the PR concerns in these sorts of posts, I believe Facebook should do more. They have so much rich information, it's incredible to consider the psychological and sociological research that could be done with it.

(I work on this team at Facebook)

There are definitely sensitive issues, but IMO the limiting factor to putting out more cool analyses is that packaging these up for public consumption takes time and effort, and there is a _lot_ on our plate for a small team. Lisa (the author) was a very smart and productive intern.

If you're qualified and want to help, drop me a line or go to facebook.com/careers

Slightly less intuitive is the fact that positive emotional updates receive fewer comments (perhaps there’s nothing more to say) whereas negative emotional updates receive more comments (perhaps as a consolation).

This is not so unexpected, given the like button. If I post a message about getting a new dog, my friends can just like it without saying anything. If I say my dog is dying, the only way to show support is by writing something. There is no sympathise or dislike button. This confirms that if you give people shortcuts for being nice, they will gladly take them.

I'm making a tool which might help your team do such analyses faster. I just sent you a message on Facebook. My contact details are also in my profile.

Not sure why, but I didn't get your message on Facebook.

That's strange. It is showing in my sent messages. Anyway, here's what I wrote:

Hey Venky,

I'm the founder of Pretty Graph (http://prettygraph.com/), a simple web-based tool for making graphs.

I noticed that your data team uses R for their analysis and graphs. Making good looking graphs can be a big time sink in R. When you are trying to publish blog posts, you want to make your graphs look nice and web-friendly.

Pretty Graph also uses R on the backend, but makes creating complicated graphs a lot easier and faster. You can import data in various formats and get a graph ready to publish in a few minutes. So your team can spend more time analysing the data rather than adjusting margins, title sizes and colours.

If it's something that interests your team, I'd love to hear from them.

Thanks, Hrishi

Feel free to reply to hrishi@prettygraph.com.

Thanks, that's cool. I'll post it on our team group.

      If you're qualified and want to help, drop me a line or go to facebook.com/careers

I'm definitely interested, and I'm currently working in Social Network Analysis in a University. Can you give us some more details on the type of stuff you guys work on?

I would email you directly, but there is no contact information on your profile.

Sorry I didn't respond earlier. We work on basically any data-analysis-intensive part of the site, including basic research into what people are using the site for, how features proliferate, deduping and disambiguation on the graph, machine learning, recommendations.

My Facebook profile is on my HN profile, and is the best way to contact me.

OkCupid tends to use the contents of private messages for their juiciest research. I suspect if Facebook started talking about what people were saying in private there'd be quite a shit storm.

Edit: Oh yeah, and OkCupid loves to make bold pronouncements on shaky data.

I'm sure those graphs are interesting, but they're so small that I can't read them, at all.

It seems that if you open them in a new window and they will be their true (larger) size.

Why can't I middle-click on them then?

My eyes hurt from trying to read them. :(

What's the deal with Tumblr, GoodReads et. al no longer being allowed in Facebook?

hmmm... are they trying to do Google (and loose their social mojo)?

Reply on news.ycombinator.com