Blog Post

Why Gravity’s Interest Graph Effort is Un-Interesting

Stay on Top of Emerging Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Earlier this week, Gravity, a Los Angeles-based startup, launched a new service called Twinterest and outlined its intent to help personalize the web through what it calls an “interest graph.” This isn’t the first time the start-up — which has garnered $10 million in funding from the likes of Redpoint Capital and August Capital — has sought the limelight by making bold (and woolly) claims.

The company, which was started by Amit Kapur, Steve Pearman and Jim Benedetto — all former Myspace (s nws) executives — came out of stealth in December 2009. Gravity initially wanted to reinvent the concept of conversations then data-mine these conversations to build interest analytics, and eventually build a business against these analytics. Suffice to say, I wasn’t a fan of the service, and took a dim view of its prospects. When it launched earlier this year in beta, the service was indeed a letdown.

Since then, the company has come up with a new strategy, and is now using Twitter feeds to build the Interest Graph, which it says it will then use to help personalize the web. Sometime in the future, Gravity CEO and co-founder Amit Kapur says, he wants to give publishers the ability to personalize content for each one of their readers.

One of the key building blocks for this “interest graph” is a new offering from Gravity called Twinterest. It’s a service that taps into the Twitter fire hose and hopes to map my interests and essentially connect them to some of the folks I might know. I tried it out and got some surprising results. For instance, the service says I have 476 interests including Country Music, Bernie Madoff, Appalachian State University, Debarge, Soviet Union and Anakin Skywalker.

Those “interests” are just straight up wrong. The rest of my “interests” remind me of a phrase we used to use often when growing up: Looking London, Going Tokyo.  What’s even funnier is that the service suggests that I should friend Rapleaf on Twitter (ironic, considering some of my writings about that San Francisco-based company). Gravity wants us to help fine-tune this “interest” data, so it can personalize the web in the future, but the Twinterest results so far make quite an un-interesting graph.

The inaccuracies in the interests displayed by Twinterest are symptomatic of some of the problems associated with natural-language processing services, which try to extrapolate my interest level in topics from proper nouns mentioned in tweets. I remember tweeting that I was listening to Debarge on the radio and feeling nostalgic. That doesn’t mean I’m interested in them.

Many services that propose to make sense of the Twitter feed and draw inferences have to deal with a whole can of worms. For starters, any service that proposes to build an interest graph or some sort of ranking, needs to do the following:

  • Analyze the content shared by a user.
  • Analyze the content of a user’s tweets.
  • Analyze what others a user follows as a potential signal of interest.
  • Analyze who follows a user to get a sense of authority of the user on that interest.

Gravity (and other services similar to them) have to figure our ways to do the aforementioned four things at scale, then build a taste graph for every single Twitter user. That’s not easy, nor cheap. One assumes Gravity has been able to do that — and have put some of their venture dollars to good use. But then as more data points are added to the mix, say Facebook, the complexity (and costs) of Gravity’s Interest Graph is only going to go up.

I think the challenge with services that use Twitter to draw inferences about me is that they only have an incomplete picture of me. My Twitter identity is very tech-centric. It ignores some of my real-world interests, and frankly, I don’t care to share a few things with the rest of the planet. In other words, Twitter can’t really help mirror the real me.

That said, it’s simple enough to build a content-discovery system based on the interest graph, and predictably, that’s why the company is building a news discovery service, The Orbit. I don’t think it’s really a very valuable service – i.e., it’s a terrible business idea. The idea of an interest graph is something that was initially championed by Hunch, which instead of doing natural-language processing, decided to start with a combination of machine learning and statistical learning.

Eventually someone will figure out a way to mine the web and build a personalized web experience. Gravity seems pretty far from it. This is something the guys who own the data – Facebook, Twitter, Google – will likely do.

Related content from GigaOM Pro (sub req’d):

Post and thumbnail images courtesy of Flickr user Luis Markovic

15 Responses to “Why Gravity’s Interest Graph Effort is Un-Interesting”

  1. Christopher

    The problem you are seeing is prevalent in most systems that utilize your “interest graph”, that is they don’t go beyond bag-of-word models. There’s much more that needs (and is) being done to sniff out your “actual interest graph” vs. “an interest graph based on a bunch of words”.

    Christopher. :)

  2. What if the best way to get users true interests dynamically within the right context…is just to get them tell it directly and explicitly !

    Twitter is a form of personalized web…iI explicitly follow people of interest and at the moment they start to spam me …i just unfollow. That is dynamic.

    Search engines use explicitly input keywords to define your interest…without even knowing users address, age, and all sort of big data used previously by marketers.

    I mean “big data” is probably the best answer to this problem. And even it was…over 90% of people interests are not share online but in their heads.

    Just a service that directly triggers my needs in very specific areas (not all), very subtly ….instead of mining it from external sources.

  3. Maganbhai Bharawad

    || The idea of an interest graph is something that was initially championed by Hunch, which instead of doing natural-language processing, decided to start with a combination of machine learning and statistical learning.

    This sentence is as vague as it can be. Anyone who deals with web content or information streams these days performs what is called statistical natural language processing which includes machine learning which is statistical predominantly in the post-Chomskyian era.

  4. “a personalized web experience”

    I don’t WANT a personalized web experience. I come to the internet to learn about other things not to stay in my cozy cocoon. I actually don’t know anybody who wants a ‘personalized experience’ following them all over the web and I work in tech (software developer since 1977, own tech business with highly tech engineering customers, worked on data presentation and user interfaces since the 70s). I mix with techs and have tech-aware kids (they grew up with computers and geeks) who have tech-aware friends. The kids are tuned out. It’s all ‘so what’. Sometimes I feel this place is an echo chamber.

  5. Hi Om:

    With Path & this post, you are back to writing the incisive blog that has always made GigaOm compelling. As other readers said, its not whether you are going to be proved wrong. Its that you are putting a clear, logical critique of an idea.

    I disagree with your simplistic list though. The scale of data is no more a problem because of Hadoop, AWS and other big data infrastructure. But a unique algorithm – a la PageRank – is the problem. Somewhere, there is a PhD student who will start a thesis, abandon the PhD and build the next TwitterRank. PageRank exploited a seemingly simple metric of a web page’s popularity rank. Somebody will have to come up with a similar metric for the social and real time graphs.

  6. Rohan Jayasekera

    What struck me most about my Twinterest results was that the things that I was criticizing or joking about were considered just as important as the things that I spoke of positively. Sentiment analysis would have been helpful in trimming the list of 1439 interests down to something more representative. As you say, the content needs to be analyzed.

  7. Shah Ullah

    You can come here for opinions that have substance; even more so when they are brutally honest. Om was wrong about Hulu and he apologizes for that, but the occasional false call doesn’t keep him or his kick-ass team from being brutally honest and anti-fanboy when it is warranted – whether its about Gravity, Path, or others. This is especially difficult to do when there are silicon valley hotshots behind a venture; there may be a perceived risk that not being a total fanboy will somehow limit your access to stories. To consider that risk and consider it a cost of doing GOOD business and reporting SUBSTANTIVE opinions is simply respectable. As a reader, I am comforted knowing that I am not reading an editor’s attempt to generate good will. I <3 GigaOM.

  8. John Doerr says, it’s all about execution.
    Me thinks, execution without deep understanding creates fluff.
    “Analyze” is easy to say, hard to do. Or we can just do another pixel sorter.

  9. Google, Facebook and Twitter will buy companies, technology and people to drive innovation. Gravity, Hunch, others focused on personalization could become extremely valuable if they get it right.

    I personally found Twinterest pretty interesting. It made me think more about what Gravity might able to do rather than what they can’t or aren’t doing.

    I’d be shocked if Gravity isn’t able to look at not only what you say, but also what you share and who you follow. Your list seems pretty obvious, and if you can do one, you can do them all. Doing it at scale, as you mentioned, is the real challenge, but Benedetto is credible from what I’ve heard.

    I find your personal observations about Twitter to be fairly trite. First, you are tech-centric. Second, most people aren’t like you. They use Twitter to vocalize more than their thoughts about their own profession.

    I’m rooting for these guys. Pessimism doesn’t have a place in innovation, only in journalism.

    • Todd

      As always you are entitled to your opinions, as I am to my own.

      On the topic of Twitter and me: you missed the point. The point is that we have a certain limited stuff we share on Twitter and as a result we end up with a special interest persona. Doesn’t matter what sector you are involved in, you are have a “centric” view, which may or may not be reflective of your true interests.

      There are a whole score of companies which are doing interesting stuff around personalization — including some like Hunch who are building their own interest graph based on data they collect instead of sifting through Twitter firehouse. I find their approach more intriguing.

      Thanks for the comment and your thoughts.