Stanford University
Information Access
Reproduced with permission from NeXT Computer, Inc.
A Reference Guide to NeXT in Higher Education, Fall 1992
ยช 1992 NeXT Computer, Inc
Intelligent text searching with the Searcher
Using Searcher, a NeXT-based text-searching program, Stanford University faculty and students can perform scholarly research and analyze text in ways never before possible. Searcher can list all the words used in an electronic document; find the number of occurrences; create a graph of the frequency of occurrences; and show, save, or print the located words in their surrounding context.
To create Searcher, George Drapeau and Jayson Adams of Stanford's Academic Information Resources used Interface Builder to design an interface for the PAT text-search engine. "NeXT let us assemble a friendly, easy-to-use interface that uses English-like commands," says Drapeau.
Users can quickly perform word or phrase searches as well as more complex searches of thousands of pages of material, including articles; research papers; numerous news files; and on-line books such as the Bible, all of Shakespeare's works, an entire Sherlock Holmes collection, and the 21,000-page Oxford English Dictionary.
"Say, for example, you wanted to find all the occurrences of the theme of love in The Adventures of Sherlock Holmes," Drapeau says. "You should be able to search for 'love' as well as for its synonyms. You should also be able to do Boolean searches-to search for 'love' and 'death' but not 'hate' within a story. And you should be able to do proximity searches-searching for all occurrences of 'murder' and 'dagger' that appear within 50 characters of one another.
"The idea here is that you couldn't do this analysis just by opening a book, even if you were to read it 10 times. With a computer, you can. By being able to analyze the way text is organized, the words an author chose to use, and where certain words and phrases appear in a text, scholars can formulate and test their theories about the work."
Faculty and students use Searcher for courses and research projects in English, philosophy, medicine, German, and other humanities and social science disciplines.
Pathology Professor Don Regula converted old pathology records into digital form and created case studies for his students from these records. Using Searcher, students studying a particular case can easily find related cases by searching by a particular symptom or doing a Boolean search for a group of symptoms.
"Before we had this tool," says Regula, "we had to synthesize a search in our own minds, give it to a computer person, and then wait to receive the results. If the results weren't quite right, we had to repeat the entire process. With Searcher, we formulate a search we think is reasonable and try it. If we can improve it-by removing a particular synonym, for example, because it brings back too many hits-we type some new characters and have another search under way in a matter of seconds."
English Professor Mary Wack has students in her undergraduate course on Chaucer's Canterbury Tales use Searcher when composing papers. "Students are not familiar with the irregular spellings and various word meanings of 14th-century English," says Wack. "This really kept them from appreciating Chaucer as a poet. The papers they wrote using Searcher were the best first set of Chaucer papers I'd ever received. What distinguished them were theses supported by literary evidence rooted in the text. Students were starting with the text itself, not with their assumptions."
Philosophy Professor Rachel Cohon started to use Searcher when she was puzzled by an argument David Hume had made in his 709-page A Treatise of Human Nature. To understand Hume's reasoning, Cohon had to comprehend how he used certain ambiguous terms in other sections of the Treatise. Once Cohon found the terms with Searcher and compared the usages, Hume's reasoning became clearer.
"Because Searcher looks through the whole volume," says Cohon, "I found some terms in places I never would have thought to look-sections of the Treatise on completely different topics. Using Searcher actually improved my thinking. I thought the changes Searcher would bring to my work would be quantitative-more speed, greater efficiency. But they were actually qualitative changes; Searcher brought an extra dimension to my research."
All in all, says Drapeau, "The NeXT machine is an effective way to deal with large texts. Its speed makes it useful for performing complex and time-consuming searches of large bodies of text. Because of the large screen space available on the NeXT, there is no clutter. And its unlimited memory allows students and instructors to obtain and review very large quantities of material."
For more information, please contact:
George Drapeau
Workstation Environments Specialist
Stanford University
Sweet Hall, Room 300
Stanford, CA 94305-3090
(415) 725-4168
drapeau@jessica.stanford.edu