How It Works

Just for You is on one level a pretty simple service. When a MyBlogLog visitor visits a site with the plug-in running, the system looks at the tags that user provided for their interests when they created their MyBlogLog account. It then finds blog posts with the same categories or tags and serves them up in a widget.

Right: MyBlogLog's Ian Kennedy is using the plug-in on his blog and it knows I like RSS. Oh yes I do.

The most interesting part, though, is it recommends posts with different but similar tags as well. "The weighting for a user's interests ranges from 10, a direct match between your posts' tags/categories and their MyBlogLog tags to 1, a loose coupling," the MyBlogLog team says. The company wouldn't tell us how exactly it does this analysis, but the algorithm was built by two members of the Yahoo! Bangalore team, Mani Kumar and Saurabh Sahni. It's very cool.

When non-MyBlogLog visitors come to your site, they will be shown recommendations based on tags from the most recent MyBlogLog users who have been there. That sounds like a good solution.

We asked the MyBlogLog team if they were going to offer further recommendations based on tags found in other accounts users have associated with their MyBlogLog profile, like Delicious or Last.fm. They said they had some top secret magic in the works for that. We'd also love to see a plug-in for other blogging platforms, like MovableType for us here at RWW.

MyBlogLog - an Incredibly Important API

MyBlogLog is a fascinating service. People expose an incredible amount of personal information about themselves in exchange for being able to see the faces of people who read their blogs. It's big outside of tech circles, too.

When the service launched its API in January, we said it was going to be a big deal. When Yahoo's Kent Brewster built the BlogJuice widget, we found functionality that we continue to use almost every day here at ReadWriteWeb. Why, we wonder, are all the coolest things built on the MyBlogLog API being built in house at Yahoo! and not by other members of the development community at large? The company says they don't know, if developers have thoughts we'd love to hear them.

Just For You is a great example of what this API can do. There are countless companies that have raised millions in venture capital to offer publishers recommendation systems for their readers - commercial publishers pay big money for this functionality. Now bloggers can have the same type of thing for free and base recommendations on the self-identified interests of their readers. That's really powerful.

We continue to be impressed by what this team is able to do with user data. This is just the kind of thing that gets us really excited about the web.