Search results

A local area network (LAN) connects computers, printers, modems and other devices located near each other, often in an office environment. The School of Library and Information Science (SLIS) at the University of Wisconsin‐Milwaukee provides a case study of selecting and using a LAN in an academic program environment. Consideration of various types of LANs took place during the selection of a LAN for SLIS. The advantages of having a LAN at SLIS have been the sharing of printers and other devices, the use of electronic mail, improvement in office management and cooperative research, and easier access to information and files available in the school.

A study of 100 series of articles in science and technology examined how well members of a series cite each other. Citations from one article to another were either excellent, good, fair, poor, or nonexistent, depending on how much information could be gotten about the cited article from the citing article. It was found that 11% of all possible citations from earlier to later articles in a series—foward citations—were excellent, good, or fair. 73% of citations from later to earlier articles—retrospective citations—were excellent, good, or fair. It is recommended that the author of a series clearly point readers to at least one retrospective member of the series and that readers use a four‐step method to identify all members of a given series.

In response to questions by Buxton and Meadows, there was an examination of the occurrence of title words in the abstracts, first paragraphs, last paragraphs and cited titles of research papers in chemistry, economics, history, mathematics and philosophy for the 1960 and 1970 eras. Title word occurrence in first paragraphs varied little among disciplines. Last paragraphs tended to have most frequent occurrence of title words in history and philosophy, and cited titles had most frequent occurrence in chemistry and mathematics. There was no significant difference between chemistry and mathematics of occurrence in abstracts; abstracts were not available for the other disciplines. Among disciplines taken as a whole, the best reflection of title word occurrence was the collection of abstracts, followed in order by first paragraphs, last paragraphs and cited titles. First and last paragraphs together provided 70% to 80% of the title words. For most disciplines, longer than average titles did demonstrate a higher frequency of title word occurrence in first and last paragraphs than did titles in general. The results implied that indexing based on extraction of title words could employ similar procedures from discipline to discipline. Nevertheless, sensitive information retrieval systems should be prepared for changes in the vocabulary of fields like history and philosophy to occur possibly more slowly than in fields like mathematics and chemistry.

The study examined the rates of use of descriptors in the ERIC system during 1966–1986 to determine if certain levels of terms were used more than others and if patterns of use were similar among hierarchies in the ERIC Thesaurus. The postings per document measure indicated how often a term had been assigned to documents during its life. This was averaged for each level in the 252 multilevel hierarchies. With little exception there was not much variation in postings per document among levels nor among hierarchies. The major exception was the mean rate of 725 postings per 100,000 documents for the broadest terms in the twenty‐nine hierarchies having four levels each. This rate was significantly higher than for the narrowest levels in these hierarchies. The lack of variation in most hierarchies suggests that all terms currently in the system are important and used by indexers. Searchers should be aware of the power of the broadest terms.

With the continuing increases in computer processing and storage capabilities, the barriers to and benefits of electronic access to more information content are becoming serious issues in information science research. The experiments described in this article, which address the value of content‐enriched access, are important to continued progress in information retrieval. Well‐selected content components and full‐text materials in electronic systems must be linked with improved search methodologies, better computer interfaces, and greater understanding of the structure and use of knowledge. Content‐enriched records, augmented by these other developments, will enhance the probability of users identifying the information they require.

The University of Illnois Information Retrieval Research Laboratory contracted with the United States Federal Emergency Management Agency (FEMA) to identify and analyze word‐oriented databases of potential relevance to FEMA. A subject profile technique was used to measure how many potentially relevant citations were found in selected databases, thus allowing a ranking and comparison of databases for the multidisciplinary field of emergency management. “Distribution of Citations in Databases in a Multidisciplinary Field” describes the ranking of databases relevant to emergency management and demonstrates the applicability of Bradford's law of scatter to citations in databases. This article describes an experiment to compare the subject profile technique used in the FEMA project to another common database coverage evaluation technique — the ‘bibliography’ or ‘review article’ technique. Although the two techniques have slightly different purposes, they can both be used to compare the coverage of databases in a particular subject area. This study shows the subject profile technique to be less costly and less time consuming.

The proliferation and popularity of microcomputers have precipitated a revolution in online searching. Complex, costly online systems such as DIALOG and BRS, normally utilized in libraries or information centers by professional searchers, have produced user‐friendly and inexpensive offspring. These scaled‐down systems, designed for use at home or in the office, are advertised and discussed in popular magazines.

The purpose of this study was to determine the rate of self‐citation in the library and information science literature. A sample of 1,058 articles was examined. 50% of the articles examined contained at least one self‐citation. Articles that were reports of research, that were written by a faculty member, that addressed a theoretical topic, or that had multiple authors were all more likely to have to higher self‐citation rates. The self‐citation rate of 50% was higher than that reported in studies of self‐citation rates in the sciences and social sciences. However, the percentage of self‐citations as related to total citations of 6.6% falls between the percentage reported in the sciences and that reported in other social sciences.

Word‐oriented databases of potential relevance to the multidisciplinary field of emergency management were identified by the University of Illinois, Information Retrieval Research Laboratory under contract to the United States Federal Emergency Management Agency (FEMA). This article is an extension and outgrowth of that contract. It analyzes forty databases for relevance to emergency management by searching each database using an emergency management subject profile, printing a random sample of citations to determine percent of false drops, and ranking the databases according to number of relevant citations. Bradford's law of scatter is shown to apply to this multidisciplinary field, using databases instead of journals and citations instead of articles. No one database provides more than 19% of the literature, however, illustrating that the literature in the field is widely scattered throughout databases. These findings can help in the choice of the specific databases containing emergency management citations and in the determination of how many databases need to be searched in order to retrieve a given percentage of the literature. A companion article in this issue of Online Review — ‘Evaluation of database coverage: a comparison of two methodologies,’ explains the subject profile evaluation method employed in this project and compares it to another coverage evaluation technique.