We select all subjects and all databases the old-fashioned way: by hand. The main goal of the Search Broker is to provide higher quality more focused search than indiscriminate search engines (for the price of less coverage).

Each search engine is tested for coverage, quality, generality, and interest before we decide to include it. If there are several search engines covering the same subject, we try to pick the best. The assignments of subjects, related subjects, and topics are done manually, and we try to put as much thought as possible into this process. We strive to be fair brokers. Our job is to bring the user our knowledge and experience with search facilities, and make it available in the most convenient way. Of course, we cannot be responsible for the content (or sometimes lack thereof), or reliability of any of the search engines.

We have not asked for permission to forward our queries to any of the search engines. We believe that we provide them with a valuable service by sending clients their way (and showing their ads). Not to mention that many of them use search software (e.g., Glimpse and Harvest) that we have developed and distributed for free. We bring the results of all searches verbatim (the only exception being the background of the page). Nevertheless, if anyone complains, we will remove any search engine promptly from our list. We do not use any service that requires registration, even for free, because we do not want to bypass the registration process. This is a research project. It is not a commercial effort.

If you want us to add your search engine, please use the Feedback Form and send us its URL with a brief description.

The Search Broker was conceived and developed by Udi Manber and Peter A. Bigot at the Department of Computer Science of the University of Arizona.
Copyright © 1997, Arizona Board of Regents

This site does not eat nor serve any cookies. Let's keep surfing low fat.