Title-Only Entries Retrieved by Use of Truncated Search Keys

Abstract

An experiment testing utility of truncated search keys as inquiry terms in an on-line system was performed on a file of 16,792 title-only bibliographic entries. Use of a 3,3 key yields eight or fewer entries 99.0% of the time.A previous paper ( 1) established that truncated derived search keys are efficient in retrieval of entries from a name-title catalog. This paper reports a similar investigation into the retrieval efficiency of truncated keys for extracting entries from an on-line, title-only catalog; it is assumed that entries retrieved would be displayed on an interactive terminal.Earlier work by Ruecking (2), Nugent (3), Kilgour (4), Dolby (5), Coe ( 6), and Newman and Buchinski ( 7) were investigations of search keys designed to retrieve bibliographic entries from magnetic tape files. The earlier paper in this series and the present paper investigate retrieval from on-line files in an interactive environment. Similarly, the work of Rothrock ( 8) inquired into the efficacy of derived truncated search keys for retrieving telephone directory entries from an on-line file.Since the appearance of the previous paper, the Ohio State University Libraries have developed and activated a remote catalog access and circulation control system employing a truncated derived search key similar to those described in the earlier paper. However, OSU adopted a 4,5 key consisting of the first four characters of the main entry and the first five characters of the title excluding initial articles and a few other nonsignificant words. Whereas the OSU system treats the name and title as a continuous string of characters, the experiments reported in this and the previous paper deal only with the first word in the name and title, articles always being excluded.

Online access will reduce author/title catalog search failure

“…Montague (1967), Richard P. Palmer (1972), James Kirkelas (1972), F. W. Lancaster (1977), and Ruth Hafter (1979) have greatly facilitated the evaluation of the literature and selection of materials to be analyzed. Also, reviewers have consistently given high rank to Sidney L. Jackson's Catalog Use Study (1958), Ben-Ami Lipetz's User Requirements in Identifying Desired Works in a Large Library (1970), 'Access and Recognition: From Users' Data to Catalogue Entries' by Renata and Richard P. Palmer's Computerizing the Card Catalogue in the University Library (1972).…”

“…As already mentioned, this analysis examines retrieval by single keyword, implied Boolean, and derived truncated search keys having a basic 3,1,1,1 formula for title (Long & Kilgour 1972). Table 4 correlates 1588 card catalog title searches with these three computerized search systems and shows that none of the computerized searches significantly improves upon user success with the card catalog.…”

“…Table 5, derived from Table 14 in the Tagliacozzo article (1970), presents a significant reduction in author-title search failures through use of computerized searches. Single keyword and derived truncated key (3,3) searches (Kilgour et al 1971) share equal success since each reduces failure by 61.3%; implied Boolean successes are 56.5%. The derived key search is to be preferred, for it will produce many fewer entries in a reply.…”

A Truncated Search Key Title Index

“…A previous paper has shown that truncated three-letter search keys derived from the first two words of a title are less specific than authortitle keys ( 1). Earlier work had revealed that addition of only the first letter of another word in a title improved specificity ( 2) .…”

Analysis of Search Key Retrieval on a Large Bibliographic File

“…A number of papers are available in the literature concerning search key effectiveness; however, all of these were done on relatively small data bases (1)(2)(3)(4)(5) . Of particular importance to this paper is Kilgour's article on truncated search keys ( 6) .…”