Entries Tagged 'Social Graph' ↓

Worth your attention: DataPortability.org - Share and remix data using open standards. Scoble has been publicly testing the portability of Facebook data (relevant posts: Facebook disabled my account, Unintended advantages, and What I was using…) in the past day and (of course) found the results to be less than ideal. I have a feeling that data portability is going to be one of the important issues of 2008. Finally.

In my feed reader just now I noticed a brief mention by John Battelle: Paid Links, Selling Links… Not Good. When you click through to the article and to the Google help page - and remembering the commotion a couple of weeks ago about some complaints that people’s Page Rank had suddenly dropped - a very encouraging pattern is starting to emerge.

Google has always taken the soundness of their systems very seriously, but I don’t think it’s coincidental that Google seems to be addressing some of the issues surrounding paid links and the like more seriously now that Facebook has made some aggressive strides into the advertising world. I think someone at Google realizes that alongside and within its ranking and presentation via search of the whole web, they also - without Orkut or OpenSocial or anything else - already have a “social graph” embedded in their databases - and one that has already proven to be more valuable than closed social networks’ social graphs.

The initial promise and reality of blogrolls (say, pre-2003), after all, was that they served as a way to declare, publicly, that such-and-such a blogger was someone you either knew or respected personally. That is the social graph right there, and Google’s always had it. The best part? There was a cost to adding someone to your blogroll (time, dilution, etc.), which served (somewhat) to pare down those lists and make them more accurate representations of bloggers’ personal preference.

The important thing about the social graph is that to be valid and useful as a commercial endeavour, connections must accurately reflect a person’s authentic relationship, whether that be with a friend, an issue-related BOF, a colleague, or anyone else. To date this has been the strength of Facebook - they made it easy for people to add friends, but through the News Feed (among other things) added a cost to doing so - which has (so far) tended to “purify” people’s contact lists in a way that MySpace’s and others’ lists never were.

Facebook is winning (by some measures) because users’ networks more closely resemble real-life relationships - Facebook isn’t, by-and-large, a friend-adding contest. Anything Google can do to ensure their results are accurate and reflect authentic relationships is likely as important in the long run as anything they do with OpenSocial. (Now if only they would do something about all of the spam blogs on Blogger).

could be explicitly communicated within the links themselves, using a schema called XFN, the XHTML Friends Network. If this sort of thing is really going to work, it’ll have to be pretty ubiquitous - things like Game Neverending and pretty much every website or application through which relationships can develop would ideally support this (or some other) scheme.

Many have been writing about Weblogs and power laws in recent days.

There’s a problem though. The articles all seem to be built on an unsupported assumption: that linking to someone is a reliable and meaningful indicator of the reading habits of the link-from weblog. No such correlation has been shown to exist.

I don’t find any support for this assumption. There are two kinds of links on weblogs: links intrinsic to the text of the site, and links in lists of “favourite sites”. When weblogs first really got going, making a links list on your own site was one of the only ways available to help yourself remember to go to the sites you preferred, so it may once have been an indicator of something. In 2003, however, there are many alternative methods that can be employed to get to the sites one wants to read. I use BlogTracker myself, but there are at least a half-dozen other ways of linking to often-read sites. My own links list (over there on the right) is partially driven by my desire to read certain sites, but links also get there because I’m polite, or because I want to reciprocate for someone who would consider my reciprocation (or lack thereof) significant, or old friends who nevertheless I don’t read often, etc.

Regarding intrinsic links, I think that there are so many different reasons why someone would link to another site - and these motivations change over time - that no causality should be inferred. Overall, though, what I would like to know is this: of heavily-trafficked sites, what proportion of their traffic comes from links from other weblogs? And, if that proportion is low, what do numbers of incoming links have to do with anything?