Search results

Searchers often avoid EMBASE due to its high cost compared to MEDLINE. However, this paper demonstrates that when searching for a variety of pharmaceutical information, retrieval from EMBASE is superior to that from MEDLINE. Duplicates were detected, but more unique citations were found in EMBASE than in MEDLINE. An EMBASE searcher has the powerful ability to use specific drug identifiers as well as brand and manufacturer names. The EMBASE record contains additional information about the drug in question that may be utilised for further searching in EMBASE. Although EMBASE's display costs are high when compared to MEDLINE, title displays are free and searching alone is not expensive. Use of DIALOG Web was found to keep the overall costs to a minimum. Employment of these cost savings measures will help the searcher to decide whether the additional drug information gained from the EMBASE record outweighs the elevated cost of its display.

A comparison of the coverage provided by BIOSIS Previews (BP) and Medline databases was carried out by VOGIN, a group of professional searchers based in The Netherlands. The area chosen was clinical medicine, and the group looked at problems requiring diagnosis and treatment. The data provided by this study show that while it is more economical to use Medline, BP provides a greater degree of completeness and often provides an unexpectedly large amount of new information. A strong point of BP is its coverage of abstracts of books and meetings, which are either not provided or difficult to find in Medline.

The paper presents an overview of the National Library of Medicine's 1988 field test and evaluation project, including the participation of 21 user sites, on the use of MEDLINE products on CD‐ROM, and reports on the subsequent evaluation forum held at the NLM on 23 September 1988.

Two questionnaires were developed in order to gain some insight into medical librarians' attitude towards the various forms in which the Medline database is offered — printed version, online database and CD‐ROM. One was sent to about 200 medical libraries in The Netherlands and the other was distributed among the users of the Core Medline/Ebsco CD‐ROM disc in The Netherlands. Responses to the first questionnaire produced answers about the altitude among Dutch librarians towards this new technology and the current rate of acceptance. Responses to the second questionnaire produced insight on the user audience, as far as current use of printed bibliographies and online services is concerned as well as the users altitude towards Medline on CD‐ROM and the preferred time span and journal selection for the CD‐ROM version of Medline.

The object of this study is to develop methods for automatically annotating the argumentative role of sentences in scientific abstracts. Working from Medline abstracts, sentences were classified into four major argumentative roles: objective, method, result, and conclusion. The idea is that, if the role of each sentence can be marked up, then these metadata can be used during information retrieval to seek particular types of information such as novelty, conclusions, methodologies, aims/goals of a scientific piece of work.

Two approaches were tested: linguistic cues and positional heuristics. Linguistic cues are lexico‐syntactic patterns modelled as regular expressions implemented in a linguistic parser. Positional heuristics make use of the relative position of a sentence in the abstract to deduce its argumentative class.

The experiments showed that positional heuristics attained a much higher degree of accuracy on Medline abstracts with an F‐score of 64 per cent, whereas the linguistic cues only attained an F‐score of 12 per cent. This is mostly because sentences from different argumentative roles are not always announced by surface linguistic cues.

A limitation to the study was the inability to test other methods to perform this task such as machine learning techniques which have been reported to perform better on Medline abstracts. Also, to compare the results of the study with earlier studies using Medline abstracts, the different argumentative roles present in Medline had to be mapped on to four major argumentative roles. This may have favourably biased the performance of the sentence classification by positional heuristics.

To the best of one's knowledge, this study presents the first instance of evaluating linguistic cues and positional heuristics on the same corpus.

The connect hour prices for use of 320 online databases up on seven major biblio‐graphic systems in February 1981 in the U.S. and Canada were analyzed. A distribution was run and all databases fell into five ranges with 65% in the medium‐low ($25–49) and medium ($50–74) ranges. Government databases were analyzed as a subset of the total, and National Library of Medicine databases were analyzed in more detail as a subset of govern‐ment online databases. Further analyses of database prices were done in terms of four new measures: (1) price per connect‐hour per record accessible online; (2) price per connect‐hour per record searchable online; (3) number of records accessible per dollar per connect‐hour; and (4) number of records searchable per dollar per connect‐hour. Data are given in both real and constant dollars. Using these four measures, one can see that database size correlates extremely well with price. Reasons for increasing the price per connect hour for MEDLINE are given. They relate to the numerical analyses and to the responsibilities government database producers have to not‐for‐profit database producers.

PaperChase is a Medline searching system designed for use by physicians and other health care professionals and students. It offers searching by subject headings, title words, author, journal title, year of publication, all of which can be combined using the Boolean operators AND, OR, and NOT. Unlike most Medline searching systems, PaperChase attempts to guide the user in developing a successful search strategy.

To assess the currency of descriptions found in SilverPlatter's Medline CDROM edition precisely, all the descriptions from 1993 were studied. The objectives were not only to quantify the mean currency of the articles, but also to establish whether the priority level, the periodicity and the subset had any influence on currency. As a secondary result, a database summarises characteristics and currency values for each serial The results show that the currency of a description can vary tremendously. Priority level, periodicity and prior database were found to have a low but highly significant influence on currency (p < 0.0001). The mean currency was 6.92 months, when Priority 1 journal articles had a mean currency of 4.53 months and appeared on CDROM two to three months earlier than other priority‐assigned journal articles. 40.28% of Priority 1 journals appeared on CDROM within three months of publication. The 10 most important serials (regarding ISI's impact factor for general medicine) are shown as examples of the database for secondary results.

CD‐ROM is an increasingly successful consumer product. The market is at present going through an experimental period with new openings being sought for specialised products, particularly in the business and financial sectors. The development of CD‐ROM products will undoubtedly be market led, becoming integrated into an end user's working environment rather than being set aside as a separate resource. Consideration of how to train end users to perform successful searches on CD‐ROM databases will become increasingly important as the market for such products grows.

This paper reviews the trend of purchasing of a printed index (Index Medicus) within one regional health authority. It shows that as the machine readable counterpart (Medline) has become more easily accessible the purchase and use of Index Medicus has increased rather than diminished.