IRList Digest Sunday, 28 August 1988 Volume 4 : Issue 46 Today's Topics: Query - Interactive video systems Reply - Pedagogical models (see V4 #39): some references - Pedagogical models: work in Netherlands Abstracts - Dissertations selected by S. Humphrey [Part 3 of 5] News addresses are Internet: fox@fox.cs.vt.edu or fox%fox.cs.vt.edu@dcssvx.cc.vt.edu BITNET: foxea@vtcc1.bitnet (replaces foxea@vtvax3) ---------------------------------------------------------------------- Date: THU, AUG 11 1988 12:33:18 From: Paul F Lambert Subject: IRList subscription ... I have a faculty member in Modern Languages who is interested in the use of Interactive Video systems for teaching. If you or any list contributors can point me to recent mailings under this topic (or other mailing-lists) I would appreciate the help. Thank you. Paul Lambert Plambert@Bentley (InfoRep) (617) 891-3467 ------------------------------ Date: Sun, 07 Aug 88 16:45 PDT From: IIN4CLB@UCLAMVS.BITNET Subject: IRlist submission about pedagogical models Ed - for IRLIST in response to Avigail Oren (C37@TAUNOS) about pedagogical models for IRS There are several literature reviews in the literature that would be helpful starting points. See my "why are online catalogs hard to use? Lessons learned from information retrieval studies" in JASIS, nov 86 (about 150 cites). Also Carol Fenichel's lit review in 1981 in library research. The last full review just on training for IR was Judy Wanger's ARIST (annual review of info sci and technology) chapter in 1979. All of these have long bibs with pointers elsewher in the literature. Hope this helps. Christine Borgman IIN4CLB@UCLAMVS ------------------------------ Date: Tue, 16 Aug 88 14:57:31 +0100 From: Frances Brazier Subject: information retrieval In reply to a request by the Computer in Education Research Lab (address unknown) on a pedagogical model for IRS, the following. Information retrieval and information retrieval strategies have been addressed within a number of projects within our Department for over 10 years now. At the moment we are building a system within which help and support are embedded to assist the process of information retrieval. This project is a collaberation between the Department of Geography (contents experts), the Department of Computer Science and the Department of Cognitive Psychology. Although this particular project has been our particular focus of activity for the last 1.5 to 2 years nothing yet has been published on the results. Results of previous projects within this field can be found in the following publications: Beishuizen, J.J. & Brazier, F.M.T. Search strategies in internal and external memories, Proceedings of the Third European Conference on Cognitive Ergonomics, Paris 1986. Brazier, F.M.T. & Beishuizen, J.J. The influence of the interface on information retrieval, Proceedings of the Fifith International Conference on Technology and Education, Edinburgh, 1988. Brazier, F.M.T. & Beishuizen, J.J. Can an interface influence information retrieval? Paper presented at the AERA Conference, New Orleans, 1988. Brazier, F.M.T. & Trimp, H.C. Teaching information retrieval skills, In Lovis, F. (ed) Proceedings of the IFIP TC3 European Conference on Computers in Education - ECCE'88 Lausanne, Switzerland, 1988. Beishuizen, J.J. (in press) Search strategies in internal and external memories, In: Human-Computer Interaction, Psychonomic Aspects, Eds. G.C. van der Veer & G. Mulder, Springer Verlag, Heidelberg Beishuizen, J.J. 1987 CIR: A computer coach for information retrieval In: Eurit 86, Developments in Educational Software and Courseware, Proceedings of the First European Conference on Eductaion and Information Technology, Eds. J. Moonen & T Plomp, Pergamon Press, Exeter Brazier, F.M.T. & Beishuizen, J.J. Information retrieval: the user interface, SIGCHI-Bulletin April 1988. I am acquainted with Marcia Bates work, but not with Penny Daniels. I too would be grateful for any references to related work and results of the project mentioned. Frances Brazier Dept of Cognitive Psychology Vrije Universiteit de Boelelaan 1111 1081 Amsterdam the Netherlands email: frances@psy.vu.nl ------------------------------ Date: Wed, 3 Aug 88 13:36:58 EDT From: "Susanne M. HUMPHREY" Subject: dissertation abstracts [Note: Part 3 of 5 - Ed.] .[ AN This item is not available from University Microfilms International ADG05-62168. AU WEDDELL, GRANT EDWIN. IN University of Toronto (Canada) Ph.D 1987. TI PHYSICAL DESIGN AND QUERY COMPILATION FOR A SEMANTIC DATA MODEL. DE Computer Science. AB In this thesis, we consider some of the problems of physical design for the more recently proposed data models. These newer models, called semantic data models, adopt concepts developed by artificial intelligence researchers investigating the general problem of knowledge representation. Our results apply to a particular choice of model, called LDM, that is also developed in the thesis. LDM incorporates the most common features of other semantic data models including a capability for a generalization hierarchy that supports multiple inheritance, support for many-valued properties and a non-procedural query language. This has the advantage that implementors of these other models can then apply our techniques for physical design to solve similar implementation problems. The performance issues we address are based on the assumption that all encoding of information is memory resident. With this assumption, some problems, such as the choice of representation for entities and simple property values, become important issues. Other issues relating to access strategies for implementing queries or to the choice of index types and their selection, are fundamentally changed. The assumption also permits us to ignore clustering problems (problems concerning the judicious placement of data in order to reduce retrieval overhead), since they then have much less relative significance to overall performance. The problems that are considered include: finding representations of entities and single valued properties, selecting a set of indices to support access to groups of entities occurring as class extensions or as values of many-valued properties, mapping transactions to forms that automatically maintain indices, and compiling queries. .] .[ AN This item is not available from University Microfilms International ADG05-61648. AU YANG, HONGZHI. IN University of Waterloo (Canada) Ph.D 1987. TI QUERY TRANSFORMATION. DE Computer Science. AB Consider a database containing not only base relations but also stored, derived relations (also called "materialized views" or "concrete views".) Relation fragments in distributed databases, database snapshots and intermediate results obtained during the processing of queries are all examples of stored derived relations. For a query, expressed in terms of base relations, we want to know whether we can answer the query by using only the data in a given set of derived relations (the set may include some base relations). The interest in the approach stems from the fact that query evaluation based on a set of derived relations may be more efficient than the traditional query processing. If the query can be represented by an equivalent expression involving derived relations, it can then be evaluated from the set of derived relations. The real problem concerns the transformation of the original query, expressed in terms of the base relations, to an equivalent one expressed in terms of the derived relations. In this thesis, we investigate the query transformation problem under the assumption that both queries and derived relations are defined by PSJ-expressions, that is, relational algebra expressions involving only projections, selections and joins. We have developed a mechanism for solving this problem. It constructs an equivalent query consisting of a set of subqueries. Each subquery can easily be computed from derived relations. A prototype of the mechanism has been implemented and some data obtained from running the prototype are also given. Query transformation is an indispensable part of a performance-oriented database system. With this problem solved, it is possible to adapt the physical database to the query load and improve the performance of query processing. .] .[ AN University Microfilms Order Number ADG88-00563. AU ZAPOROZHETZ, LAURENE ELIZABETH. IN University of Oregon Ph.D 1987, 165 pages. TI THE DISSERTATION LITERATURE REVIEW: HOW FACULTY ADVISORS PREPARE THEIR DOCTORAL CANDIDATES. DE Education, Higher. AB Thirty three active doctoral advisors on a College of Education faculty in a research oriented university participated in the study. They filled out questionnaires and participated in lengthy interviews in which they described how they advised their doctoral candidates on the literature review portion of the dissertation. The overall response pattern from the 33 advisors indicated the following beliefs and general advising procedures. They rated (1) refereed journals, (2) books, (3) dissertations, and (4) ERIC, as the most productive bibliographic formats for the dissertation literature review in education. They ranked the literature review chapter the lowest of the five traditional dissertation chapters when they reflected on their level of advising expertise, and on the amount of time/energy they gave to a chapter. They ranked the research/methodology chapter highest on both counts. Some advisors, especially senior faculty, expressed distrust and little knowledge of computerized literature searching technologies. They expected their advisees to have bibliographic skills at the doctoral level, even if the skills are not generally taught in graduate programs. Most advisors indicated they advised the way they were advised, reporting that they were left on their own to learn bibliographic skills during their graduate years. Overall, they felt doctoral candidates should go to the library to "do the literature review", come back with the results, and at that stage the advisor should offer suggestions and assist in writing and editing. The findings suggest that graduate programs should (1) consider bibliographic instruction on a par with research methodology instruction in the preparation of doctoral candidates, and (2) update faculty advisors on new searching technologies in order to increase their advising effectiveness. New technologies for accessing bibliographic data, and the generally low level of advising for the dissertation literature review suggest many areas of further research. .] .[ AN University Microfilms Order Number ADG88-02888. AU SHOFF, HARRY L. IN Indiana University of Pennsylvania Ph.D 1987, 208 pages. TI MATERIALS AND STRATEGIES FOR LITERARY RESEARCH IN THE COMPUTER AGE. DE Education, Language and Literature. AB The study of literature is not exempt from the exponential growth of information. The growth affects literary research to the point that the scholar must reassess traditional research sources as well as learn the new information technology. This study explores the traditional sources of information in language and literature and changing utilization strategies. It also assesses the direct impact automated information storage and retrieval devices and techniques have had on the research process. The project is designed to form the core of a bibliographic methods text. Most studies of the influence of computers on research neglect the impact information growth has had on research materials that are not yet computerized. Paper format sources are still important, but the strategies for using them have substantially altered. The first part of this research examines traditional sources of information. The chapters assess each of the primary areas of research such as subject headings, classifications, biography, periodicals, reference sources, government documents, and bibliographies. The discussions in each category delineate the characteristics of the standard research sources. In addition, strategies are developed which permit the researcher to plan a more efficient inquiry. The argument addreses the impact information growth has had on standard sources such as the recent trend for consolidated indexes that cover dozens or hundreds of individual volumes. The second part of the study treats the direct effect automated information storage has had on literary research. Computerized storage of information primarily affects the literary researcher because of three major developments. These developments are: (1) Libraries are replacing traditional catalog listings with automated catalogs and computer access. (2) Libraries are combining their catalog holdings with many other libraries to form computerized union catalogs. (3) Specialized bibliographies in the various disciplines are combining with hundreds of other bibliographies to form large bibliographic databases. The final chapters of the dissertation explore the characteristics of these computerized aids to research and analyze future trends. .] .[ AN University Microfilms Order Number ADG87-28256. AU AKERS, JUANITA SMITH. IN Southern Illinois University at Carbondale Ph.D 1986, 134 pages. TI BEFORE AND AFTER: IMPACT OF THE LOSS OF THE ILLINOIS RESOURCES AND DISSEMINATION NETWORK. DE Information Science. AB The purpose of this study was to discover and describe in more detail the final impact, if any, of computer-generated information dissemination in Illinois after federal funding was discontinued. The descriptive method was chosen for this study. The data sources for this discovery were factual information and the perceptions from state department liaisons who delivered the information product and provided other forms of technical assistance, and school district administrators, who were identified as users of the information service. To achieve the purpose of this study, a questionnaire was developed and mailed to the two groups identified for participation in this study. The total population consisted of twenty state department liaisons and 80 school district administrators. Of the 100 surveyed, 96.0 percent of the respondents returned questionnaires. The researcher recommended that further study be conducted to determine more effective methods of disseminating the research product and of controlling the quality and cost of computer-generated research. It was recommended that once an information dissemination project shows clearly and dramatically that it makes a desirable difference, every effort should be made to institutionalize that program or its successful element for the period of time needed to enhance school improvement. Furthermore, it was recommended that during the process of making application for funding, incentives be awarded those agencies who provide a realistic and economical management plan for achieving institutionalization. Finally, it was recommended that monies presently allocated for school improvement activities be used, in part, for computer generated information dissemination in Illinois school districts. The researcher concluded that there is an above average need for computer-generated annotated bibiographies within the public elementary and secondary school districts in Illinois. Positive steps should be taken to reestablish the information service for public school educators. This action should be the direct result of commitment and resolve on the part of the funding agency(s) and the agency(s) administering the program. In addition to reestablishing the program, other initiatives should be taken to improve the methods of disseminating the educational information, and the quality of the information product disseminated. .] .[ AN University Microfilms Order Number ADG88-02457. AU CHAUNCEY, CAROLE AGATHA. IN Case Western University Ph.D 1987, 252 pages. TI AN ASSESSMENT OF INDEX TERM RANKING. DE Information Science. AB The enhancement of document retrieval is dependent on the use of index terms which convey adequate information. It has been postulated that index terms are not equally content-bearing and therefore cannot promote the retrieval of relevant documents with the same degree of dependability. This investigation assesses how several indexers rank the selected index terms, as well as how the agreement/consistency patterns among indexers vary within the titles and the abstracts. The hypotheses tested in this study are that (a) terms selected during an indexing exercise have varying degrees of agreement; (b) the most content-bearing terms get assigned earliest in the indexing exercise; (c) high agreement terms are found in specific parts of the text. To test these hypotheses the following tests were conducted; (a) exhaustivity-time experiment (Productivity test); and (b) inter-indexer consistency versus exhaustivity. The results of the exhaustivity-time test show (a) an increase in selection time as exhaustivity increases; and (b) inter-indexer consistency decreases as exhaustivity increases. .] .[ AN University Microfilms Order Number ADG88-00197. AU COTTRILL, CHARLOTTE ALICE. IN The Union for Experimenting Colleges and Universities Ph.D 1987, 176 pages. TI A CO-CITATION STUDY OF THE SCIENTIFIC LITERATURE OF TWO INNOVATION RESEARCH TRADITIONS: DIFFUSION OF INNOVATIONS AND TECHNOLOGY TRANSFER. DE Information Science. AB The state of a research field can be studied through the publications produced by its members. The most revealing data from the literature of a field are the references cited by its practitioners in their publications. These references to earlier literature provide information about the field as well as the authors. References can be studied with co-citation analysis which is a type of network analysis that identifies the relationship among either cited documents or the combined works of individual scholars. A co-citation occurs when two documents (or authors) are cited together by a third and later document. The strength of the relationship is the number of times the pair are cited together by later documents. This relationship is established by the citing authors, and is therefore an unobtrusive measure of the "field's view" of the relatedness of two works which is based on the collective judgements of many citers. Co-citation analysis has been used to study many research fields and shown to reflect with considerable accuracy the membership of a specialty and its content and problem orientation. The present study is a descriptive structural analysis of the diffusion of innovations and technology transfer literature from 1966 through 1972. These two research traditions represent groups of scholars within the larger innovation research scientific community. Using samples of documents drawn from the ISI Social Science Citation Index, the documents were reduced to groups of works by authors, and a modified author co-citation analysis was performed to identify the relationships among the works of these scholars. Separate analyses were performed on the diffusion of innovations and technology transfer co-citations and the combined groups of co-citations. Cluster analysis was used to identify groups of highly similar works. These clusters of similar works represent subspecialty groups of scholars within each of the two areas and the combined area. Multidimensional scaling was used to generate maps spatially representing the relationships among the subspecialties in subject space. As publications, particularly in journals, constitute the formal communication system of science, the present study is essentially a study of scientific communication. The analyses employed are network methods that were used to investigate communication flows and identify types of relational patterns. .] [Note: continued in next issue - Ed] ------------------------------ END OF IRList Digest ********************