From: Northwestern University
Watson solves mystery of searching information on the Web
EVANSTON, Ill. --- Sherlock Holmes had his Watson, Alexander Graham Bell had his Watson, and now, thanks to Northwestern University's Intelligent Information Laboratory (The InfoLab), all researchers can have a Watson as well.
Watson, developed by InfoLab graduate student Jay Budzik, is a personal information management system that can turn your PC into a research librarian by collecting information relevant to your project.
"The key to Watson is that, while you're writing, the system actually reads the documents you are working on and uses its understanding of your subject matter to search the Web for the pages and sites most relevant to your work," said Kristian Hammond, InfoLab director and professor of computer science in the Robert R. McCormick School of Engineering and Applied Science. The system then sifts through the pages to remove dated material and redundancies, thus providing useful information while reducing the volume of results.
"Whenever you do searches on the Web, you often get hundreds of thousands of responses to a query," Hammond said. But, by posing its queries in the context of the project you are working on, Watson manages to cut through the volume of material to the information you need.
Watson accomplishes that by working in tandem with your Web browser and your PC's word processing program. "Watson bridges the gap between researchers and the information they need," Hammond said. "By watching as you work, it can actually figure out your needs and build queries to online systems that are richer and more precise than any but those that the most sophisticated user can construct. The beauty is that the system doesn't require any work on the part of the user. You just do your job and Watson does its."
Whether your research project is a report for your business or a doctoral dissertation, Watson can "watch" you work and shape its responses to the type of project you are working on. It mines not only the Internet, but also other internal and proprietary sources, continually making suggestions as you work. Watson's search includes not only text, but also photos and graphics to enhance your research project.
"Watson spends most of its time working in the background," Hammond said. "But you also can use it to talk with search engines directly. The difference is that when you type a query into Watson it already knows the subject you are working on and only looks for relevant documents. For example, if you are writing a paper on construction equipment and you ask Watson to find documents related to 'Caterpillar,' it will never return pages related to fuzzy insects."
Technically, Watson is a Java application that sits on your computer screen while you are working on a project and serves as a go-between from your browser or word processing program to the World Wide Web. It is available free by downloading the beta version of the program from the InfoLab's Web site at http://infolab.nwu.edu/watson. The final version will be available once the testing is evaluated.
Once it's downloaded onto your PC, Watson links itself to Internet Explorer and begins its detective work, doing searches for you as you work. (A version of Watson that works with Netscape will be available soon.)
Northwestern University's InfoLab is a collection of computer science faculty and graduate students who are working to develop the technology to form a bridge between people and the machines that serve them. The InfoLab is driven by the desire to not only develop new ideas but also to facilitate the transfer of these ideas into working systems.
Contact: Kristian Hammond