Information Retrieval List Digest 128 (September 8, 1992) URL = http://hegel.lib.ncsu.edu/stacks/serials/irld/irld-128 ========================================================================= Date: Tue, 8 Sep 1992 17:10:32 PST Reply-To: "Information Retrieval List" Sender: "Information Retrieval List" From: IRLIST Subject: IR-L Digest, Vol.IX,No.32, Issue 128 IRLIST Digest ISSN 1064-6965 September 8, 1992 Volume IX, Number 32 Issue 128 ********************************************************** I. NOTICES A. Meeting Announcements/Calls for Papers 1. Classification Society of North America, Annual Meeting, June 24-26, 1993, Pittsburgh, PA B. Publications Announcements 1. RLG's Citadel Service 2. Netbook C. Miscellaneous 1. IR-L Policy 1. IR-L Policy II. QUERIES B. Requests for Information 1. Creating MS-DOS Hypertext CD-ROM? IV. PROJECT WORK C. Abstracts 1. IR-Related Dissertation Abstracts: Correction ********************************************************** I. NOTICES I.A.1. Fr: Stephen Hirtle Re: Classification Society of North America Classification Society of North America Call for Papers 1993 Annual Meeting Twenty-fifth Anniversary of the Society June 24-26, 1993 Pittsburgh, PA The 1993 annual meeting of the CSNA will be held on the campus of the University of Pittsburgh, starting Thursday evening, June 24 and continuing until Saturday afternoon, June 26. The purpose of the annual meeting is to provide a forum for presentation and cross-disciplinary discussion of data-oriented approaches related to classification and understanding of complex data structures. This meeting marks the twenty-fifth anniversary of the CSNA. Invited sessions will feature the application areas of Numerical Taxonomy, Numerical Ecology, Molecular Biology, and Information Retrieval. The program will also include plenary talks by invited speakers and symposia on Neural Network, Genetic Algorithm, Multivariate Density Estimation and Visual Clustering approaches to classification. As in the past, there will be invited and contributed papers by CSNA members, non-members and students in the areas of cluster analysis, multidimensional scaling and related methods of exploratory data analysis and their use in applications. The programs for previous meetings can be found in the Journal of Classification. Because of the informal nature of the meeting, speakers often present research that is currently in the meeting, speakers often present research that is currently in progress. The meeting will be preceded by optional short courses on Classification and Clustering and on Multivariate Density Estimation and Visual Clustering on Thursday, June 24. We are interested in soliciting presentations in all areas of the scientific study of classification and clustering. Persons wishing to present a talk, propose a symposium, or request information about local arrangements, should contact Stephen C. Hirtle Department of Information Science University of Pittsburgh Pittsburgh, PA 15260, U.S.A. tel: (412) 624-9434 fax: (412) 624-5231 email: csna93@lis.pitt.edu Paper submissions must be on a single page and include the following: the title of the talk; the name, institutional affiliation, mailing address and electronic mail address (if available) of each author; an abstract of not more than 200 words; and a list of a few keywords to help classify the paper. Please identify the presenter in the case of multiple authors. If possible, each submission should be classified according to one of following areas: applications, classification, clustering, exploratory data analysis, graph theory, graphics and visualization, latent class models, multidimensional scaling, network models, numerical taxonomy, pattern recognition, statistical methods, or other (please specify). Abstracts will be due March 15, 1993. Submission via electronic mail is encouraged. Contributed papers will be allotted 20 minutes, including discussion, unless otherwise approved by the Program Committee. Abstracts will be distributed to conference attendees, but no Abstracts will be distributed to conference attendees, but no other proceedings will be published. Some papers may be published later in the Journal of Classification. Herman P. Friedman, president of CSNA, will be the co- chair of the program committee along with Stephen C. Hirtle. Local support for the meeting will be provided by the School of Library and Information Science at the University of Pittsburgh. ********** I.B.1. Fr: RLIN Info. Ctr.--PDC Re: RLG's Citadel Service (The following is an announcement from the Research Libraries Group.) FREE 30-DAY TRIALS FOR RLG's CITADEL SERVICE Free 30-day trials of CitaDel's citation files are now available to interested institutions. The files include: ABI/INFORM, Periodical Abstracts, Newspaper Abstracts, Dissertation Abstracts, Ei Page One, PAIS '80+, History of Technology Bibliography, Index to Foreign Legal Periodicals, Index to Hispanic Legislation (World Law Index, pt. 1), and Avery Index to Architectural Periodicals. THE CITADEL SERVICE: CitaDel is the new article-citation and document-delivery service from the Research Libraries Group (RLG). Through CitaDel, libraries, universities, and other institutions can make citation databases available to their users over a library or campuswide network without having to load and maintain large files locally. To subscribe to CitaDel files, institutions simply pay a fixed annual fee for the file (or files) of their choice, and their users receive unlimited searching free of charge. DOCUMENT DELIVERY: Full-text copies of articles cited in ABI/INFORM, Periodical Abstracts, Newspaper Abstracts, and Ei Page One can be ordered online through CitaDel. Document-delivery options include first class mail, fax, and Ariel (RLG's document-transmission system for the Internet). FREE TRIAL SUBSCRIPTIONS: Institutions may try one CitaDel file or several for 30 days without charge. There is no obligation to purchase a subscription after a 30-day trial. (Limit per institution: One free 30-day trial per file; however, several CitaDel files may be tried concurrently.) COMING SOON TO CITADEL: * More files, including Hispanic American Periodical Index, Isis History of Science Bibliography, and the INION Bibliographies (from the Russian Institute of Scientific Information in the Social Sciences). * Pricing by the search (for institutions not desiring subscription pricing). * Document delivery for more files. For more information about CitaDel and free 30-day trials, please contact: The RLIN Information Center, 1200 Villa Street, Mountain View, CA 94041-1100 USA; e-mail: BL.RIC@RLG.BITNET or BL.RIC@RLG.STANFORD.EDU (Internet); phone:1-800-537-RLIN (US & Canada); FAX: 415-964-0943 ********** I.B.2. Fr: Bazil Farid Re: NetBook - a Network-based Information Retrieval System I am posting this for a friend - please do not send e-mail to this account. please call (713) 789-1746 for any questions or ordering information - thanks - Bazil ANNOUNCING NETBOOK... Tera Knowledge Systems, Inc. a Houston based software firm, announces NetBook, a network-based Information Retrieval system. It is capable of searching textual databases and other collections on remote machines located across networks that are connected via TCP/IP. NetBook conforms to the ANSI Z39.50 protocol. It comes with an OSF/Motif graphical user interface (the client) and a retrieval server. NetBook's indexing and search engines are based on data structures that use Optimal Minimal Perfect Hashing (OMPH) technology. NetBook features advanced searching and relevance feedback methods that provide high precision. NetBook's indexes full text in various formats such as text files, Unix mail, Gnu rmail, Netnews format, PICT files, GIF files and TIFF files, Internet resource guide, postscript files and other mail formats. NetBook indexes most text at a speed of about 300 Mb per hour. NetBook is also capable of displaying various formats such as text, PICT files, TIF and GIF files. NetBook currently runs under UNIX (SunOS, Interactive UNIX) and a Windows 3.1 version will be announced shortly. For full information call (713) 789-1746 or write Tera Knowledge Systems, Inc. P.O. Box 771862 Houston, TX 77215-1862 ********** I.C.1. Fr: N. Gusack, IR-L Moderator Re: IR-L Policy: More Comments 1: Steve Cisler I think it's fine to continue posting various announcements, even if they are exclusive of some groups. That, in itself, is an education for many of the readers, and may serve to work against the organizers of such meetings. I think that printing the complaints is good too. I learned from them too. Imagine, Morris dancers in Riyadh! Steve Cisler sac@apple.com ********** 2: Friedrich Gebhardt Ladies and gentlemen, While I am strongly opposed to denying people the attendance of meetings on grounds of their sex, religion, nationality or similar reasons, I am just as strongly opposed to being told by an editor which conferences should be avoided on such grounds. Nobody is forced to attend a conference just because it is listed somewhere (e.g. in IR-LIST). In case a policy of suppressing conferences in Saudi Arabia is adopted, then please also suppress all conferences in the US since the US denies visas to persons having AIDS. Friedrich Gebhardt, Bonn. E-mail: gebhardt@gmd.de ********************************************************** II. QUERIES II.B.1. Fr: Theordore Allan Re: Creating MS-DOS Hypertext CD-ROM? Need help re: currently popular/respected MS-DOS based hypertext Need help re: currently popular/respected MS-DOS based hypertext development products. We are trying to work up a project plan in a SHORT "window of opportunity" for funding--something other University folks can appreciate, I'm sure--for what will turn out to be a sort of a hypertext book on CD-ROM. A professor here is already a successful print publisher, but wants to take the text of his last work and update it in electronic form this time rather than just in print form. This means the text, figures, and tables, and a way to get to each. For example, when the text says, "in Figure 12 you can see the results of..." you have a command, button, or function-key to press which will "pop up" or otherwise activate "Figure 12". I'm not defining whether "Figure 12" will be a scanned image or redrawn artwork, or whether it will be .PIC, .PCX, .TIF, .EPS, or redrawn artwork, or whether it will be .PIC, .PCX, .TIF, .EPS, or whatever--I suspect I'll be more limited by the products which can show this and text, than by how I can deliver the Figures. It is a reference work, so the "chapters" follow a regular format. I can see the opportunity to cross "data base" and "hypertext/linking" lines here, since the text content organization is so regular. I've been hanging around the edges since Hypertext '89 in Pittsburgh, but this will be my first real -funded- project (we hope!). Otherwise, I haven't worked much with anything but HyperCard (and a -little- look at Owl Guide). This project is defined as being MS-DOS based. I'm inclined to try to stay away from Windows, from a market standpoint as well as for complexity's sake, but that's by no means a given. So, what are currently popular MS-DOS based products that would lend themselves to creation of an MS-DOS CD-ROM like this? Will something like MaxThink be able to handle the text links (like to tables or other articles which are related) and the "figures"? Will I have to use Windows-based products, or are there non-Windows products which can fill the bill? Thanks! Theodore Allan (Ted) Morris, University of Cincinnati, Medical Center Information and Communications, Information Retrieval and Development, 231 Bethesda Ave., ML#574, Cincinnati, OH 45267-0574 513-558-0177V, 558-0758F, MORRISTA@UC.EDU, MORRIS@UCUNIX.SAN.UC.EDU, NTS WB8VNV. Call me up and I'll talk data to ya! ********************************************************** IV. PROJECT WORK Issue 127 contained a major error in this section. Lines were copied from the wrong file and the first and last abstracts are incomplete. IV.C.1. Fr: Susanne M. Humphrey Re: Selected IR-Related Dissertation Abstracts The following are citations selected by title and abstract as being related to Information Retrieval (IR), resulting from a computer search, using BRS Information Technologies, of the Dissertation Abstracts Online database produced by University Microfilms International (UMI). Included are UMI order number, title, author, degree, year, institution; number of pages, one or more Dissertation Abstracts International (DAI) subject descriptors chosen by the author, and abstract. Unless otherwise specified, paper or microform copies of dissertations may be ordered from University Microfilms International, Dissertation Copies, Post Office Box 1764, Ann Arbor, MI 48106; telephone for U.S. (except Michigan, Hawaii, Alaska): 1-800-521-3042, for Canada: 1-800-268-6090. Price lists and other ordering and shipping information are in the introduction to the published shipping information are in the introduction to the published DAI. An alternate source for copies is sometimes provided. Dissertation titles and abstracts contained here are published with permission of University Microfilms International, publishers of Dissertation Abstracts International (copyright by University Microfilms International), and may not be reproduced without their prior permission. AN University Microfilms Order Number ADG92-05887. AU BYRER, JOYCE KAY. TI THE RELATIVE EFFECTIVENESS OF A DIRECT MANIPULATION INTERFACE VERSUS COMMAND-BASED INTERFACE IN A DATABASE QUERY TASK ENVIRONMENT. IN Indiana University Ph.D. 1991, 281 pages. SO DAI V52(09), SecA, pp3334. DE Business Administration, General. Psychology, General. Computer Science. AB Over the past decade, five different styles of interaction between users and computers have evolved: (1) menu selection, (2) form fill-in, (3) command language, (4) natural language, and (5) direct manipulation. All five have proven successful for particular applications. However, one of the five styles, direct manipulation, stands out as the most "visual". A major question posed by both practitioners and software developers is "Do direct manipulation interfaces provide a more effective interface for computer users?" Researchers have examined particular aspects of the diversity of interaction styles and techniques and have determined that direct manipulation interfaces have not been shown to be universally more effective. Human-computer interface studies have identified a complex relationship between the computer task, users who carry out the task, the environment in which they work, and the information system that they are using. This study investigated one particular combination of these factors that has received little attention to date. More specifically, this study compared novice users performing database query tasks with two direct manipulation interfaces and a command-based interface. Fifty-one subjects were randomly assigned to one of three database interfaces. During each experimental session, each subject was given "hands-on" training with a computer in the construction of database queries. As the subjects proceeded through the training, they were given three sets of query tasks, each successively more difficult. Comparisons focused on the effects of the different database query interfaces on user performance (as measured by the correctness in tasks and performance time) across the different database tasks (one-table, two-table, and three-table), as well as on user attitudes (as measured by confidence in their answers and perceived ease-of-use). The results indicated that direct manipulation interfaces, as a whole, do not lead to significantly better user performance in formulating table queries. Users of one of the direct manipulation interfaces performed very similarly to the users of the command-based interface. On the other hand, users of the second direct manipulation interface performed significantly better than users of the other two interfaces. This study suggests that the design features that minimize user errors may be a more important criteria in determining user performance than whether the interface is direct manipulation or command-based. AN University Microfilms Order Number ADG92-01555. AU WU, BRUCE JIINPO. TI THE EFFECTS OF DATA MODELS AND CONCEPTUAL MODELS OF THE STRUCTURED QUERY LANGUAGE ON THE TASK OF QUERY WRITING BY END USERS. IN University of North Texas Ph.D. 1991, 192 pages. SO DAI V52(08), SecA, pp2989. DE Business Administration, Management. Computer Science. Education, Business. AB This research is an empirical investigation of human factors on the use of database systems. The problem motivating the study is the difficulty encountered by end-users in retrieving data from a database. The independent variables are the users' mental model of the system, and the data model. The dependent variable is the user's query writing performance, as measured by the number of syntax and semantic errors, as well as the amount of time required to complete a query task. A laboratory experiment approach was employed. The experiment was structured as a three-by-three factorial design. One factor was the type of the data model. Data models were used to present database contents to end-users. The second factor was the type of the conceptual model of the structured query language. The conceptual models with different descriptions of the structured query language were used to build users' mental model of the query language system. Students were used as surrogates of business computer end-users. Test cases were developed and administered to students. Student query writing performance was coded and analyzed. The results of statistical analysis on the collected data exhibited several critical findings. The number of semantic errors committed when writing database queries was affected by the utilization of conceptual models. The diagram model as a conceptual model of the structured query language was superior to other conceptual models. The semantic errors committed in near-transfer high-complexity tasks were affected by data models. The syntax errors committed either in all query tasks or in each task category were neither affected by the use of different data models, nor influenced by the use of different conceptual models. The time required to complete query tasks was not affected by either data models or conceptual models. No interaction effects were found between the two manipulated variables. ********************************************************** IRLIST Digest is distributed from the University of California, Division of Library Automation, 300 Lakeside Drive, Oakland, CA. 94612-3550. Send subscription requests to: LISTSERV@UCCVMA.BITNET Send submissions to IRLIST to: IR-L@UCCVMA.BITNET Editorial Staff: Clifford Lynch lynch@uccmvsa.ucop.edu or calur@uccmvsa.bitnet Nancy Gusack ncgur@uccmvsa.bitnet Mary Engle meeur@uccmvsa.bitnet The IRLIST Archives will be set up for anonymous FTP, and the address will be announced in future issues. To access back issues presently, send the message INDEX IR-L to LISTSERV@UCCVMA.BITNET. To get a specific issue listed in the Index, send the message GET IR-L LOGYYMM, where YY is the year and MM is the numeric month in which the issue was mailed, to LISTSERV@UCCVMA (Bitnet) or LISTSERV@UCCVMA.UCOP.EDU. These files are not to be sold or used for commercial purposes. Contact Nancy Gusack or Mary Engle for more information on IRLIST. THE OPINIONS EXPRESSED IN IRLIST DO NOT REPRESENT THOSE OF THE EDITORS OR THE UNIVERSITY OF CALIFORNIA. AUTHORS ASSUME FULL RESPONSIBILITY FOR THE CONTENTS OF THEIR SUBMISSIONS TO IRLIST.