Internet Softbot

The Softbot was one of the 5 finalists in the 1995 Discover Awards for Technological Innovation in Computer Software.

Building autonomous agents that interact with real-world software environments such as operating systems or databases is a pragmatically convenient yet intellectually challenging substrate for AI research. To support this claim, we are utilizing planning and machine-learning techniques to develop an Internet softbot (software robot), a customizable and (moderately) intelligent assistant for Internet access. The softbot accepts goals in a high-level language, generates and executes plans to achieve these goals, and learns from its experience. The softbot enables a human user to state what he or she wants accomplished. The softbot disambiguates the request and dynamically determines how and where to satisfy it. The softbot uses a UNIX shell and the World-Wide Web to interact with a wide range of internet resources.

Take a tour of the softbot's graphical user interface.

Principal Investigators: Oren Etzioni, Daniel Weld.

Also, check out the MetaCrawler Softbot --- a fielded Web service that enables you to search multiple Web Indices in parallel, and provides sophisticated pruning options. Try it!

For further information, contact:

Oren Etzioni (etzioni@cs.washington.edu)

An accessible introduction to the Softbot project can be found here:

Technical softbot papers can be found here.

A cartoon representation of the Internet softbot taken from the L.G. Blanchard article that appeared in the December 1994 issue of Columns, the University of Washington alumni magazine.


The softbots research group is currently:


The Softbot-hackers info web (local access only)


[homepage Button] Back to AI Home Page
[CSE Button] Back to CSE Home Page
Mike Perkowitz (map@cs.washington.edu)