JASIS HomeJASIS TOC's

Journal of the Association for Information Science

IN THIS ISSUE

 

Bert R. Boyce

485

RESEARCH

 

Professional Summarizing: No Cognitive Simulation without Observation
Brigitte Endres-Niggemeyer and Elisabeth Neugebauer

Endres-Neggemeyer and Neugebauer empirically observe and isolate the work steps of the document representation process to create a model of the summarization mechanism. A toolbox of strategies used in summarizing processes is identified, and used to create a conceptual model which is projected onto a blackboard design.

 

486

 

Scholarly Communication and Electronic Journals: An Impact Study
Stephen P. Harter

Using a database of peer-reviewed electronic journals drawn from two directories, Harter did cited work searches for 39 ejournals on the three ISI citation databases. Using DIALOG to create a reasonable collection of the various forms of the ejournal titles to be found, considerable effort was devoted to a manual cleanup of the results. Eleven were published in both print and electronic formats, and citations make no distinction as to which format was utilized. Fifteen ejournals were not cited and 13 were cited between one and five times. Eight were cited ten or more times; three from the sciences and five from the social sciences. Thus overall impact must be considered low. The top ejournal had 1500 citations in 1994 alone, but it is available in both formats. The top pure ejournal has 190 citations, and PACS Review, some of whose volumes were published in book form had 111 citations. These top three, when compared to other journals in their disciplinary area, ranked high in impact factor, but poorly in total number of articles published.

 

507

 

A Comparison of Group and Individual Performance among Subject Experts and Untrained Workers at the Document Retrieval Task
W. John Wilbur

Using a cosine vector approach, the 50 closest documents to 100 query documents chosen randomly by Wilber were isolated. Judges compared each set of 50 to the query document assigning relevance on a scale. Using seven subject expert judges, six were considered a panel and the seventh the target user, with target user rotated to increase the data points. A panel of six learned, but other than subject expert judges, was also used. Groups were more effective than individuals at predicting the judgments of an unknown user, and other than subject expert groups performed nearly as well as expert groups. Because some documents are rated high by most judges and others are rated high by few or only individual judges, only by pooling can documents of wide appeal be identified. This infers the need for emulation of group rather than individual judgements in retrieval processes.

 

517

 

Citation Context versus the Frequency Counts of Citation Histories
Sinisa Maricic, Jagoda Spaventi, Leo Pavicic, and Greta Pifat-Mrzljak

Maricic et al., selected 219 cited papers and their citing paper. The physical section of the citing paper where the cited paper occurred was noted and a point value assigned. The citation's meaningful (high) or cursory (low) nature was also judged and using these two values a combined ordinal scale was created. A factor analysis seems to indicate separability by location. One factor is dominated by the introduction, one by the results section and the third by an inverse relationship between methodology and conclusion sections. The time delay of citation has little effect on this analysis. Cursory citations are found only in the introduction and meaningful citations are in all sections. The numerical analysis and the context analysis provide varying results and caution against evaluation on raw citation counts.

 

530

 

User Satisfaction with Information Seeking on the Internet
Harry Bruce

Measures of satisfaction lack clear definition and tend to be multi-variate constructs based on a combination of the results of rating scales on several variables providing only ordinal measures. Satisfaction for Bruce is measured using cross modality matching to achieve an interval scale. Two hundred e-mail addresses were invited to take part and 37 academics agreed. After being interviewed concerning past information seeking experiences on the Internet, satisfaction estimates were solicited and received in terms of numerical magnitude estimates, a one to six category rating, and in terms of grip force exerted. The measures for satisfaction have correlations similar to those for estimates of line length by the same subjects, and the ratios for line estimation and satisfaction are virtually identical. Subjects regard themselves as infrequent users, are self taught and have a high expectation of success in Internet searching. No relationship was found between Internet training and satisfaction, or between satisfaction and frequency of use.

 

541

 

Testing the Maximum Entropy Principle for Information Retrieval
Paul B. Kantor and Jung Jin Lee

Kantor and Jung Jin Lee test the Maximum Entropy Principle for retrieval using the TREC5 database and a binary classification based on the presence or absence of terms in text. MEP performs best with small data sets and progressively worsens as database size increases. Whereas it's use is certainly computationally tractable, it seems unlikely that the principle accurately reflects the distribution of terms across relevant and non-relevant texts or that it will lead to enhanced retrieval.
 

557

BOOK REVIEWS

 

The Internet Searcher's Handbook: Locating Information, People, and Software
edited by Peter Morville, Louis Rosenfeld, and Joseph Janes
Reviewed by: Valerie Jaffe
 

567

 

Proceedings of the Sixth Conference of the International Society for Scientometrics and Informetrics
edited by Bluma C. Peritz and Leo Egghe
Reviewed by: Judit Bar-Ilan
 

568

 

Information Science: Still an Emerging Discipline
edited by James G. Williams and Toni Carbo
Reviewed by: Charles H. Davis
 

569

 

Internet Economics
edited by Lee W. McKnight and Joseph P. Bailey
Reviewed by: Christinger Tomer
 

569

 

Modern Information Systems for Managers
by Hossein Bidgoli
Reviewed by: Yonathan Mizrachi
 

571

 

Information Seeking in Context
edited by Pertti Vakkari, Reijo Savolainen, and Brenda Dervin
Reviewed by: Terrence A. Brooks
 

574

 

Visual Explanations: Images and Quantities, Evidence and Narrative
by Edward R. Tufte
Reviewed by: Robert J. Skovira
 

574

 

Information Services for Secondary Schools
by Dana McDougald and Melvin Bowie
Reviewed by: Carol A. Doll
 

575

LETTERS TO THE EDITOR

   

577

ASIS HomeSearch ASISMake A Comment

© 1998 , Association for Information Science
Last update: November 06, 1998