- social software: automatic relationship clustering
 - Machine Learning Meets the User Interface, Dec 12, 2003
 - John Platt’s Home Page (MSR)
My research is focused on helping people become more effective and efficient. This research goal has two parts: making
intelligent user interfaces and building more efficient machine learning
algorithms to support the interface.Int the field of building intelligent user interfaces, I have worked on
- Improving the resolution of LCDs with ClearType.
 - Making digital music easier to use with audio fingerprinting and AutoDJ.
 - Making personal photographs more browsable with AutoAlbum.
 - Improving document search through text categorization.
 - Enabling Chinese users to quickly enter text with handwriting recognition.
 
I have also helped to create improved machine learning algorithms, such as
 - Clustering algorithms (an overview)
 - CURE: An Efficient Clustering Algorithm for Large Databases (1998)
 - Clustering Algorithms (basic basic)
 - Automatic Categorizing vs. Good Searching: Which Approach Will Ease Email Overload The Most?”
 - CRM114 possible uses/mod
 - Outlook Categorizer Add-In: An Experimental Framework for Email Categorization and Management
 - List of open source clustering algorithm implementations – including Java, Python, Perl libraries
 - dbacl – a digramic Bayesian classifier