Research Interests

My basic research interests can be clustered into a few different areas: tools and techniques for managing and learning about digital information, economics and learning in markets of digital information goods, and the social impact of technology.

Community Connections

For the past four years, I've been the director of Community Connections, a technology-based service-learning program. Service-learning is an educational philosophy that seeks to help students understand the root causes of social phenomena through community service. Our work addresses the digital divide, both in San Francisco and internationally.

We've worked in San Francisco with a number of community partners, including Network Ministries, St Anthony's Foundation, BayviewMAGIC, and Good Samaritan Family Resource Center, helping to set up and maintain labs and equipment, develop educational materials, and teach classes.

We've also conducted four trips to Tacna, Peru, where we've set up labs in two local schools, assisted an organization for working children, and taught classes for both teachers and students. Veanne Cao created an outstanding documentary about our most recent trip, which you can see here.

Managing Digital Information

In the past couple of years, I've gotten interested in developing tools to help users deal with the vast amount of information available on the World Wide Web, particularly text and hypertext data.
Slashpack
As a precursor to developing tools for filtering information or learning user preferences, we realized that it would be necessary to build a generic toolkit for collecting, storing, and processing information. We dubbed this toolkit Slashpack (Semi-Large Scale Hypertext Package). Slashpack consists of a set of Python modules for collecting and processing hypertext from a wide variety of sources, including the web, file systems, web services, and special-purpose archives, storing the data in a file system, and writing metadata to a database.

Christopher H. Brooks, Monica Agarwal, Jason Endo, Ryan King, Nancy Montanez, and Rudd Stevens. Slashpack: An Integrated Toolkit for Gathering and Filterng Hypertext Data (poster) Proceedings of the 21st National Conference on Artificial Intelligence (AAAI-06). Boston, MA, July, 2006.
Citepack
Citepack is a research paper recommendation system built on top of Slashpack. It allows users to tag papers they are interested in and crawls the web for releated resources, including papers, software, conferences, and courses. We use a support vector machine to learn a model of each user's preferences in order to provide better recommendations.

Christopher H. Brooks, Yeh Fang, Ketaki Joshi, Papanii Okai and Xia Zhou. Citepack: An Autonomous Agent for Discovering and Integrating Research Sources. Proceedings of the 2007 AAAI Workshop on Information Integration on the Web. Vancouver, BC, July, 2007.

Tagging in Blogs
Along with Nancy Montanez, I've been investigating the efficacy of tags as a mechanism for categorizing and retrieving ingformation in blogs. We've shown that tags are primarily helpful as a means for constructing broad categories, as opposed to recalling specific articles. We've also shown how the content of blog posts can be used to induce a tag taxonomy using hierarchical clustering, and used this to develop tools that suggest related tags based on the content of a post.

Christopher H. Brooks and Nancy Montanez. Improved Annotation of the Blogopshere via Autotagging and Hierarchical Clustering Proceedings of the 15th World Wide Web Conference (WWW06), Edinburgh, Scotland, May 2006.

Christopher H. Brooks and Nancy Montanez. An Analysis of the Effectiveness of Tagging in Blogs. Proceesings of the 2005 AAAI Spring Symposium on Computational Approaches to Analyzing Weblogs. Stanford, CA. March 2006.

Multiagent Learning and Information Economics

My earlier research deals with the problems that arise when computational agents in an economic setting must learn how to efficiently price information goods. These include: how to select a model of consumer preferences that efficiently trades off the fraction of available profit extracted from the consumer against the time needed to learn the model, understanding when learning producers are better off targeting niche markets, and the usefulness of niches (or congregations) as a mechanism for efficient search and coordination in large-scale multiagent systems. Please see the publications page for more information.