ISR11
Scientific Report No. ISR-11 Information Storage and Retrieval
The SMART System -- Retrieval Results and Future Plans
chapter
G. Salton
Harvard University
Gerard Salton
Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government.
I-1
I. The SMART System --
Retrieval Results and Future Plans
G. Salton
1. Introduction
The SMART system is a fully-automatic document retrieval system, capable
of processing on a 7O[OCRerr] computer search requests and documents available in
English, and of retrieving those documents most nearly similar to the
corresponding queries. The machine programs, consisting of approximately
150,000 program steps, can be used not only for language analysis and
retrieval, but also for the evaluation of search effectiveness by processing
each search request in several different ways while comparing the results
obtained in each case.
The initial emphasis in the experimental runs performed with the SMART
system was placed on the use of a large number of fully automatic language
analysis procedures, including dictionary look-up as well as statistical
and syntactic methods, and on the evaluation of the relative effectiveness
of each procedure for indexing and search purposes. At the time of
this writing, extensive experiments have been performed with four document
collections in three subject areas documentation, computer science, and
aerodynamics. Notwithstanding the apparent diversity in the subject matter
treated, the search results were found to be basically the same in each of
the three areas, in the sense that procedures which appear to operate well
in one area also exhibit a superior performance in the others. Furthermore,
a comparison of the automatic text analysis methods with the standard