When Microsoft’s Chief Software Architect Ray Ozzie announced the creation of the Future Social Experiences (FUSE) Lab last fall to focus on social technologies, it was unclear what he expected the group to work on.
Now, as the first experiments to emerge from Microsoft’s (NASDAQ: MSFT) social investigations begin to see the light of day, the types of projects the lab is going to pursue are coming into sharper focus.
Exhibit A is “Project Emporia,” a tweet filtering system for people who want to follow what’s most interesting to them among the literally millions of Twitter posts daily but don’t have a lot of technical knowledge about how the social system works.
“Project Emporia filters updates from the Twitter public feed and automatically develops lenses on topics based on the likes/dislike preference of readers, providing suggested insight into what’s happening in Twitter around that topic seen through people’s eyes,” reads a Microsoft overview document on the project.
Users can select from pre-defined “lenses” — for instance, some of them include categories such as technology, entertainment, sports, news, and business — in order to see the latest tweets in those areas.
The main screen also lets users set the level of relevance that the filters identify, ranging from none to five stars.
Although it is only a demonstration project and may never be incorporated into actual shipping Microsoft products, the FUSE Lab has made Project Emporia available for testing by interested parties as what it’s calling an “alpha release.”
In that regard, some of the functionality of the filters has been limited for now, specifically a feature that lets users refine their filtering by using a “like/dislike” feature.
“The number of individuals who can use the full features of Project Emporia will increase as we learn, adapt and scale the service,” said an FAQ on the FUSE Lab site.
The filtering mechanism is not just the result of work by the FUSE Lab. Microsoft’s basic research group — known as Microsoft Research or MSR — participated in the creation of the “matchbox” recommender system.
“Under the hood, Project Emporia uses a recommender system trained on like/dislike feedback given by users with a particular ‘lens’ on the world thereby providing insight into what’s happening in Twitter around the topic lens. This allows users find and tune topic areas they are interested in without having a Twitter account. Project Emporia also aggregates tweets which point to Web pages,” the FAQ said.
Project Emporia is located here.
Stuart J. Johnston is a contributing writer at “>Internet.com, the network for technology professionals.