Matchbox Technology offers new Personalization Capabilities

March 12th Bing announced an ambitious new approach to personalization they are calling Matchbox Technology. Matchbox is reported to be able to understand the Web more like a human would. This enables the search engine to predict what the searcher might want to see.

Traditionally, collaborative filtering systems assume that two people like the same things. However with fast breaking information like news where neither user has had a chance to view the subject, how can you predict what people will like?

Matchbox is an attempt to answer this problem in personalization by not only making predictions on what you might want to see based on other similar people's activities, but also by understanding content like a human would. The new technique called Feature Generalization is launching first on Project Emporia, a Windows Phone 7 app. and Web site that lets you experiment with the technology also on a desktop.

Project Emporia is designed to filter news stories by automatically predicted news categories. When we tested this, we got the feeling we had seen something like this before, and we have...at Pandora. The music streaming site that lets you listen to music that others like you enjoy. Like Pandora, Project Emporia lets you choose Show me more like this or Show me less like this to fine tune your preferences.

We assume after the service learns what you like to read in regards to news, it will start showing you content you'll likely be interested in. Emporia connects to both your Facebook and Twitter accounts to provide information about what your friends like - but it's unclear if they have to be an Emporia member or not.

While this particular technology is not likely to see the mainstream Bing News service in ...