The Noisy Channel

 

Google Follow Finder

April 14th, 2010 · 4 Comments · Uncategorized

I know there’s lots of interesting stuff coming out at the Chirp Twitter developer conference this week, and I’m still catching up on it all. But I am happy to point folks to a Google Labs application that was announced this morning: Follow Finder.

It’s not the first application to suggest Twitter followers based on analysis of the social graph, but I’ve actually found its suggestions to be quite plausible. For example, it suggests @fredwilson, @cshirky, @mattcutts, @peteskomoroch, and @msftresearch as “tweeps” I should follow, and suggests that the following users have similar followers to mine: @endeca, @lemire, @yahooresearch, @googleresearch, and @mattcutts.

There’s a bit of an “everything sounds like Coldplay” effect (e.g., @fredwilson shows up in a lot of the searches I tried), but overall I’m impressed with the quality, especially compared to the other suggestion tools I’ve tried.

4 responses so far ↓

  • 1 jeremy // Apr 14, 2010 at 8:27 pm

    Fun. I added 2-3 people after trying it out.

    But I found that 80% of the recommendations I got were people that I was already following. Twasn’t much use, that. Unless Google is explicitly trying to take advantage of my confirmation biases, to win me over :-)

  • 2 Tapajyoti // Apr 14, 2010 at 8:35 pm

    It would be interesting if they considered content along with the links to make the suggestions.

  • 3 Daniel Tunkelang // Apr 14, 2010 at 9:26 pm

    Yeah, kinda lame to recommend people you already follow–though not entirely useless, especially for people who don’t actually pay attention to most of the folks they follow.

    Re considering content: it would be interesting, but I’m impressed at what they infer from the link graph. Not sure that using the content would make that much of a difference to the overall suggestions. What would be interesting is to organize the suggested users based on the kind of content they produce–but I suspect that’s a much harder problem.

  • 4 Zainul Franciscus // Apr 14, 2010 at 10:18 pm

    I have been reading some Social Networking Analysis (SNA) theories and applications. One of the concept in SNA is that we can identify information sources based on the frequency and quality of the conversation between the people in the network

    If we can extend the concept of social graph to identify “who” these people converse the most, we can create a suggestion list based on the opinion leaders or information source from twitter.

Clicky Web Analytics