How is iBrarian different?

The difference in a nutshell: iBrarian is a proactive research assistant that classifies and presents scholarly papers, articles, news and blogs, not Web pages.

iBrarian is a fully automated computer system that reads and learns from scholarly papers and articles as well as online news and political blogs. To those working in the field iBrarian is a large-scale exercise in fully automated deep learning. To everyone else it is an intelligent library. It uses what it learns to present to its users an (hopefully) intelligent network of related topics with corresponding papers and articles. As it reads more it learns more and as it learns more it presents better and better information. Being a tireless computer system iBrarian is constantly reading and constantly learning more.

iBrarian is intended as an intelligent resource for those seeking scholarly references. Searching with iBrarian is done by entering a topic, not a sequence of keywords. iBrarian responds to inquiries by presenting a network of closely related topics as well as a list of subtopics based on your entry. You are free to navigate this network of topics to find one that best matches what you are looking for, or perhaps one that sparks your curiousity. When you have found the topic that best matches what you seek iBrarian can then take you to lists of papers and articles and, of course, to the actual papers and articles themselves.

iBrarian is a library of works produced by scholars, commentators and other writers. iBrarian archives and can present all of its cataloged works. Being a library it does not contain encyclopedic entries as Wikipedia does nor does it perform ranked keyword searches of Web pages as Google and other search engines do. For those types of searches please see the respective sites.

So feel free to give iBrarian a try. Enter the name of a topic (two or three words work best) and see where it takes you.

To understand more about the information presented by iBrarian please see Why are some of these results funny?

For more information on iBrarian itself please see the iBrarian Project page at

Copyright 2008
Comments about this site? email us at
Interested in iBrarian? Contact Professor Eric Berkowitz