ISR11
Scientific Report No. ISR-11 Information Storage and Retrieval
Operating Instructions for the SMART Text Processing and Document Retrieval System
chapter
M. E. Lesk
Harvard University
Gerard Salton
Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government.
II -15
EXPAND a a is either [OCRerr] DOCS, or AT.T. and specifies whether
requests only, documents only, or all input is to be
expanded.
3.[OCRerr]. Vector Ebcpansion by Means of Concept Hierarchies
In addition to statistical means for expanding document vectors, one
may wish to expand vectors by means of a pre-assigned concept hierarchy.
Various types of relationships between concepts may be specified and used
in expansion procedures. The hierarchy conBists of structured trees of
concepts, in which each concept has a clearly defined "parent'1, plus a
set of "brotherst', and `150n5t1. To designate looser connections, each te[OCRerr]
may be assigned a set of "cross-referencesT1. [OCRerr]xpansion involves the intro-
duction into the concept vector of the sons and/or brothers and/or father
and/or cross-references of each term in the original vector. The selection
of the method of expansion is by mean3 of a weighting scheme, and is
discussed in part 3.5.
To perform a hierarchical expansion the specification HIER a is
given where a is either EXPAND or SHRINK. [OCRerr]XPAND indicates that the
final vector is to contain the original vector plus the related concepts
from the hierarchy; SHRINK indicates that the initial vecto? is delet[OCRerr]d
fr[OCRerr]EL the final vector, and only the expanded concepts are used. For
exsmple, if it is desired to generalize a vector, each concept could be
replaced by its parent using HIER SHRINK; HIER [OCRerr]PAND would merely augment
each concept with its parent.
EXPAND a serves the same function as CONCO[OCRerr]; a is either RE[OCRerr],
DOCS, or AMa and indicates respectively that req[OCRerr]ests, documents, or all
input texts are to be expanded.