Microsoft Future
Social Experience Labs and Cambridge University have quietly been developing and testing what company and university engineers call a "personalized" news stream by reusing news floating through the
social Web. The app, Project Emporia, supports phones running Windows and HTML 5 devices.
Users vote on the stories with a plus or minus sign, and the app uses those votes to continually serve
up new content. While the project began as a tool for Twitter, now it acts as a personal mobile search engine running on Windows 7.
Project Emporia reuses data from social networks,
applying the data for other means. Ralf Herbrich, the principal development manager for Microsoft Bing, demonstrated the search engine earlier this month at the Next11 conference in Berlin. A description on the Next11 Web site refers to it as "computational intelligence technologies on
large online data collections." Herbrich heads the Bing personalization team that focuses on prototyping and enabling personalized experiences across Microsoft's Online Services Division.
Project Emporia relies on two technologies from Microsoft Research: auto categorization and matchbox. The
app gets all its news information from shared URLs on Twitter -- about 1 million daily -- and the combined technology works similar to the way marketers analyze the relevance of ads.
The auto
categorization allows users to personalize interests by category, such as technology or news, and allows them to set up the app to see more of any type of story. It recommends article topics based on
content, hosting site and popularity in respect to the Twitter audience.
Microsoft FUSE Labs also has been working on a project called Montage, which lets users author stories from friends or
a page of news from company news on a mobile phone. It lets people create a Web page within seconds. It also develops crowdsourced news on the page.