Effective Information Retrieval Using Term Accuracy

The performance of information retrieval systems
can be evaluated in a number of different 
ways.  Much of the published evaluation work is based
on measuring the retrieval performance of an average 
user query.  Unfortunately, formal proofs are difficult
to construct for the average case.  In the present 
study, retrieval evaluation is based on optimizing the
performance of a specific user query.  The concept 
of query term accuracy is in troduced as the probability
of occurrence of a query term in the documents 
relevant to that query.  By relating term accuracy
to the frequency of occurrence of the term in the 
documents of a collection it is possible to give formal
proofs of the effectiveness with respect to a 
given user query of a number of automatic indexing systems
that have been used successfully in experimental 
situations.  Among these are inverse document frequency
weighting, thesaurus construction, and phrase 
generation.

CACM March, 1977

Yu, C. T.
Salton, G.

information retrieval, automatic indexing, content
analysis, term accuracy, frequency weighting, 
thesaurus and phrase transformations

3.71 3.72 3.75

CA770302 JB December 30, 1977  1:52 AM

2990	5	2990
2990	5	2990
2990	5	2990
3134	5	2990
1976	6	2990
2160	6	2990
2524	6	2990
2990	6	2990