★ wanayoo — archive 1999 http://www.cs.washington.edu/research/projects/softbots/www/projects.html

Softbots (software robots): are intelligent agents that use software tools and services on a person's behalf...

Tool use is one of the hallmarks of intelligence. In many cases, softbots rely on the same tools and utilities available to human computer users --- tools for sending mail, printing files, and so on...

Softbots allow a person to communicate what they want accomplished and then dynamically determine how and where to satisfy the persons request. Softbots developed at the University of Washington interact with a wide range of software tools and services through the World Wide Web...

The softbot family tree below provides an overview of softbot development. The red buttons link to functional softbots and the blue buttons link to information about research prototypes.

MetaCrawler

The MetaCrawler Softbot is a fielded Web service that enables you to search multiple Web Indices in parallel, and provides sophisticated pruning options. To use MetaCrawler click on the icon above. To find out more about MetaCrawler continue on to MetaCrawler Information.

recommends MetaCrawler as the search service of choice!

The Ahoy Softbot is a White Page service for the Web that locates individual's home pages with high accuracy. To use Ahoy! click on the icon above. To find out more about Ahoy! continue on to Ahoy! Information.

calls Ahoy! ``the most reliable single tool for tracking down home pages...''

The Automated Travel Assistant Softbot is a fielded Web service that enables you to find the optimal flight to suit your travel needs. To use Automated Travel Assistant click on the icon above. To find out more about Automated Travel Assistant continue on to Automated Travel Assistant Information.

Internet Softbot

The Internet Softbot Research Prototype is an assistant for Internet access that helps first-time computer users find their way around the Internet. The Internet Softbot allows you to state what you want accomplished and then goes on to find answers using a wide range of resources available on the Internet and the Web. To find out more about the Internet Softbot continue on to Internet Softbot Information.

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

Occam

The Occam Research Prototype is an information gathering engine for the Internet. Occam uses a heuristic (problem solving) planner that enables you to locate information from database queries. Occam is designed to be small, efficient and distributed, so that it can eventually reside on the Web as a java application or a desktop client. To find out more about Occam continue on to Occam Information.

Internet Learning Agent

The Internet Learning Agent Research Prototype is an information resource locator that enables you keep up with the growing number and variety of information resources available on the Internet. The Internet Learning Agent forms models of information resources by interacting with them. To find out more about Internet Learning Agent continue on to Internet Learning Agent Information.

Shopbot

The Shopbot Research Prototype was retired in 1998. It was a WWW shopping agent that enabled you you to shop for computer software or CD's by manufacturer or title and artist queries. This technology was adapted in private industry and contributed to the development of an even better shopping agent at www.jango.excite.com.

[FLIPPER]

The Flipper Softbot was retired in late 1997. Flipper was a prototype Oceanographic Collection Softbot whose task was to find and retrieve oceanographic data from the Internet.

The prototype implemented several key automation capabilities to access known Internet sites. It included an application program interface that used the CGI (common gateway interface) to allow our agent to directly access the information behind HTML pages that are normally only accessible to a person. It queried the databases automatically and in parallel (for speed), sorted the query responses by a rudimentary data quality measure, and retrieved the data to the agent. Although not very intelligent, this Oceanographic Collection Agent prototype was still able to shift several significant information management burdens away from the human.

Comments to pjallen@cs.washington.edu

November 1998