This week we launched the new Microsoft Research Hub page at research.microsoft.com/hub.
The Hub page is an experiment in creating a fully automated page which is driven by data about the latest and the most popular research from MSR.
The Hub page is dynamically generated from RSS feeds. It’s like an RSS reader web app, where RSS is the only data format and multiple RSS feeds are the only data source for the Hub page. The RSS feeds reflect primitive web lifecycle machine learning:
- MSR staff publish content to the research.microsoft.com web platform [RMC]; this content gets automatically added to the relevant feed. For example, a newly published publication will be added to the New Publications feed.
- RMC visitors read or view what MSR staff have published to the RMC web platform.
- The most read, viewed published content gets automatically added to the relevant feed. For example, the most viewed video will be added to the Popular Videos feed.
- Orthogonal perspectives from MSR staff and their colleagues who post on social networks [Facebook, Twitter, blogs] are added to Blogs and Community feeds.
- The Popular feeds are based on the visitor activity for the previous week, month, and year.
- Back-end aggregation capability creates and updates the feeds from internal and external sources.
There’s been a lot of chatter in the blogosphere and web development communities about the death of RSS. I don’t buy it. As the Microsoft Research Hub page demonstrates, RSS is a rich semantic format for creating data that can easily be consumed by web apps.