Updates (most recent on top):

• I recently moved to Los Angeles. I now work out of the Burbank Yahoo! office, but make semi-regular trips back to the Bay Area.
• Ivory v0.1 was released. Ivory is a MapReduce-based retrieval toolkit that supports various retrieval models, including the Markov Random Field model for IR. The project is a joint effort with Jimmy Lin and Tamer Elsayed at the University of Maryland.
• I finally updated my CV to reflect what I've been up to for the past two years.
• Search Engines: Information Retrieval in Practice, a book I co-authored with W. Bruce Croft and Trevor Strohman is now available! The book can be purchased online at Amazon or Barnes and Noble.
• I completed my Ph.D. during the summer of 2007. I'm currently a Research Scientist at Yahoo! Research in Santa Clara, CA. My new email address is metzler AT yahoo DASH inc DOT com.

What I did at UMass:

• During my time at UMass, I was a research assistant at the Center for Intelligent Information Retrieval (CIIR) working under Professor W. Bruce Croft. My thesis research primarily focused on statistical models and machine learning for information retrieval.
• My thesis can be downloaded here -- thesis.

Education:

• Ph.D.. in Computer Science at the University of Massachusetts, Amherst, 2007.
• M.S. in Computer Science at the University of Massachusetts, Amherst, 2005.
• B.S. in Computer Science and Mathematics, at Rose-Hulman Institute of Technology, 2002.

Projects I've worked on:

• Markov Random Field models for information retrieval -- My thesis subject was Markov random fields for information retrieval. I investigated both the theoretical side of the issue (e.g. modeling and parameter estimation), as well as the practical side of things (e.g. how we can use the model to achieve good effectiveness for various retrieval tasks).
• Searching using Markov Random Fields (SMRF) -- A Java-based toolkit being designed to support research into the Markov Random Field model for information retrieval. More details soon!
• RECAP -- a tool for exploring, analyzing, and visualizing various kinds of text reuse within a collection of documents.
• Indri -- an efficient, scalable search engine with a robust query language. Indri, a collaborative effort with Trevor Strohman and Howard Turtle, synthesizes and enhances the Lemur and INQUERY search tools. My research focused on the development of the underlying retrieval model and query language. Click here for publications, presentations, and other materials related to Indri.

Quick links:

• Indri parameter files and queries used at the TREC 2005 Terabyte Track.
• A detailed overview of Indri [ppt][pdf]
• Indri query language quick reference
• Indri retrieval model overview
• Perl script to generate dependence model queries for Indri