IRList Digest Tuesday, 11 August 1987 Volume 3 : Issue 26 Today's Topics: Abstracts - Dissertation abstracts relating to inf. retrieval (pt 3 of 4) News addresses are ARPANET: fox@vtopus.cs.vt.edu BITNET: foxea@vtvax3.bitnet CSNET: fox@vt UUCPNET: seismo!vtisr1!irlistrq ---------------------------------------------------------------------- Date: Thu, 6 Aug 87 16:45:32 EDT From: Susanne Humphrey Subject: new edition of abstracts for IRList and SIGIR Forum Selected IR-Related Dissertation Abstracts (continued - part 3 of 4) AN University Microfilms Order Number ADG87-11620. AU SACHAR, HARVINDER JIT SINGH. IN The University of Texas at Arlington Ph.D. 1986, 125 pages. TI Theoretical aspects of design of and retrieval from similarity-based relational database systems. SO DAI v48(02), SecB, pp498. DE Computer Science. AB The similarity-based relational data model originated by Buckles/Petry permits the representation of inexact information in the form of a relational database. In this dissertation, a theoretically sound definition of fuzzy functional dependency was developed. Inference rules for fuzzy functional dependencies were presented and proved to be sound and complete. The concept of a fuzzy key for similarity-based data model was developed and normal forms were defined for this data model. These developments provide the capability to incorporate the fuzzy information about the real world in the design process. To provide a higher-level query language for this data model, a complete fuzzy domain relational calculus was designed which is an extension of domain calculus for ordinary relational databases. It was proven also that for any fuzzy relational algebra expression, there is an equivalent safe formula in fuzzy domain calculus. As a first step towards more advanced query methods for the similarity-based data model, the concept of dual measure of possibility/necessity as applied to simple non-fuzzy queries for this data model was developed. AN University Microfilms Order Number ADG87-08507. AU TYLER, SHERMAN WILLIAM. IN University of Pittsburgh Ph.D. 1986, 189 pages. TI SAUCI: self-adaptive user-computer interface. SO DAI v47(12), SecB, pp4964. DE Computer Science. AB Different approaches to the design of the human-computer interface have been taken in the past. These can be organized into four broad categories: tack-on; intuitive/empirical; formal; and conversational. There are several important interface design criteria that have never been adequately attained in any of these approaches. One is modularity, that is, maintaining a clear separation between the interface and its target system. A second criterion is self-adaptation, or the ability of the interface to modify its own behavior to suit a given individual user. Two further criteria relate to the interface's potential to guide users in performing typical high-level tasks on the target system and to provide intelligent advice on the use of that system. This research was focused on developing an integrated technique for achieving these four design criteria. To that end, an abstract architecture called SAUCI, or the Self-Adaptive User-Computer Interface, was proposed, embodying a knowledge-based, object-oriented approach to interface design. The foundation of this approach rests upon information encoded within sets of objects. This information includes separate knowledge bases describing the individual users, the commands of the target system, and the high-level tasks appropriate for that system. The behavior of the interface is controlled by various methods which call upon the knowledge bases in a rule-governed manner to decide what interface features should be present at each phase of the user's dialogue with the target system. To test the feasibility of the proposed architecture, a working interface was implemented on a Xerox 1108 computer in the LOOPS language, with a UNIX operating system running on a separate minicomputer as the target system. An empirical evaluation of this prototype revealed clear advantages over the standard interface. Closer examination pointed to each of the factors of modularity, task guidance, and user-tailored assistance as playing a significant role in these effects. A discussion of additional applications of this architecture and of areas for future development is offered as further evidence of the value of this approach as a general framework for human-computer interface design. AN University Microfilms Order Number ADG87-08805. AU MORRIS, JOHN HENRY. IN Iowa State University Ph.D. 1986, 127 pages. TI An experimental interface of a microcomputer with a Vega universal testing machine to retrieve data on test specimens. SO DAI v48(01), SecA, pp111. DE Education, Technology. AB The purpose of this study was to develop an inexpensive interface between the Vega universal testing machine and a Commodore-64 microcomputer. Also, to develop a computer program that will store and retrieve pertinent information about the metallurgical properties of tensile specimens (i.e., modulus of elasticity, Brinell Hardness number, yield point, reduction of area, tempering temperature, etc.). Also, compare the interfacing with the conventional method. It is theorized that by interfacing the Vega universal testing machine with the Commodore-64 microcomputer, the metallurgical data calculated by the computer will be equal to the metallurgical data calculated by the conventional method. Out of eleven hypotheses, there were nine hypotheses with significant difference at the ninety-five percent confidence level. Two hypotheses had no differences. These were the percent elongation and unit deformation. The percent elongation is a multiple of the unit deformation by 100 times. The computer was faster overall than the conventional method. However, the regression predictive equation used in the computer program to calculate the metallurgical data produced on the average higher values than the conventional method. The predictive equation can be adjusted to calculate values that are equivalent to the correct values for any given specimen. AN This item is not available from University Microfilms International ADG03-74461. AU BOETTCHER, KEVIN LLOYD. IN Massachusetts Institute of Technology Ph.D. 1986. TI A methodology for the analysis and design of human information-processing organizations. SO ADD X1986. DE Engineering, Electronics and Electrical. AN University Microfilms Order Number ADG87-12104. AU DURANCE, PAUL WILLIAM. IN The University of Michigan Ph.D. 1987, 373 pages. TI Application to logical design to incomplete medical record processing. SO DAI v48(02), SecB, PP532. DE Engineering, Industrial. Health Sciences, Hospital Management. AB Current hospital-based information systems are not adequate to meet the rapidly expanding information needs generated by additional economic competition and payment system changes. Hospitals face both increasingly aggressive competition from other institutions and substitution of treatment modalities which reduces demand for inpatient services. Simultaneously, cost minimization incentives have been established by the change from cost-based reimbursement to prospective fixed price and capitation payment systems. Hospital managements need cost effective information processing as well as more information to manage their institution effectively. Since Medical Record Departments (MRDs) currently provide most medical information within hospitals, interest is strong in expanding information processing capabilities, while also increasing processing efficiency. In particular, completing the inpatient medical record is a slow and costly process that does not satisfactorily meet post-discharge information demands. An automated information processing system (IPS) that uses the latest in document image storage and retrieval technology is proposed. This IPS for aiding in inpatient medical record processing is referred to as the Record Completion System (RCS). Software Engineering principles are applied to the determination and specification of some RCS requirements. In particular, a Logical Design methodology (LDM) is developed for the creation, capture, and analysis of logical functional requirements. The LDM combines Structured Systems Analysis and Petri Net theory. The Entity Relationship Attribute (ERA) model that defines a logical design language within the LDM is automated at the Program for Research in Information Systems Engineering (PRISE) and applied in the hospital-based requirements specification project. A database that represents the Logical Design Specification (LDS) for the RCS application is populated and reviewed. The discussion of the proposed RCS reveals that a number of current problems, which result in slow and inadequate incomplete medical record processing, can be alleviated or eliminated. It is concluded that: (1) an IPS based on document image storage and retrieval should improve both medical information availability and completion processing; (2) further work on the development of such an RCS is warranted; and (3) continued development and application of the LDM should be pursued. AN This item is not available from University Microfilms International ADG03-75232. AU LOSEE, ROBERT MACLEAN, JR. IN The University of Chicago Ph.D. 1985. TI The performance of probabilistic models of document retrieval systems. SO ADD X1986. DE Information Science. AN University Microfilms Order Number ADG87-11733. AU MONTGOMERY, JACQUELINE DELORES. IN The Florida State University Ph.D. 1987, 256 pages. TI End user searchers of online bibliographic databases in an industrial setting: training, use and perceived benefits. SO DAI v48(02), SecA, pp238. DE Information Science. AB A study of potential end users of online bibliographic searches was conducted in three parts at a product development location of a consumer products company. Part one was a survey of the population. Part two included collection and analysis of search data from a sample trained to conduct online bibliographic searches in either a one-day course or a two-hour seminar. Part three consisted of interviews with the sample at the close of the data collection phase. The survey data showed that interest in learning to search was associated with highest degree earned and type of previous exposure to online searching principles. The search data could not be analyzed for statistical significance of differences between the two types of training due to uneven search activity by subjects. Approximately half of the subjects trained did not conduct searches during the nine-month study, and of the searches obtained, more than half were conducted by only two subjects. Interview data revealed that availability of intermediaries and alternative resources, the perception that needed materials would not be available online, and difficulties in using computers and telecommunications software contributed to information users' decision not to conduct online bibliographic searches. Searchers considered convenience, time saved, and convincing others to support a decision as positive attributes of online searches. Concerns about frustration in use and missed information were viewed as negative attributes. Nonsearchers considered online searches conducted for them to have the positive attribute of convenience, and the negative attributes of time used and cost. Searchers and nonsearchers considered non-online sources to have positive attributes of convenience, time saved, relevance, and expertise, and negative attributes of information overload, time used, concerns about correctness and frustration in use. This research may be used in management decision-making regarding feasible approaches to training end users to conduct their own searches, or to communicate with intermediary searchers. AN University Microfilms Order Number ADG87-06331. AU QUARLES, PATRICIA O'HARA. IN New York University Ph.D. 1986, 211 pages. TI Videotex and conceptions of knowledge: an historically-based model for identifying epistemological consequences of videotex. SO DAI v47(12), SecA, pp4217. DE Information Science. AB The purpose of this study was to develop a model for evaluating the potential epistemological consequences of the videotex medium. It was theorized that an analysis of critical attributes of the medium would yield a series of possible consequences. In order to develop the list of critical attributes and potential consequences, the work of five prominent media historians and theorists was examined. Eric Havelock, Walter Ong, Elizabeth Eisenstein, William Ivins and Harold Innis were the core group from whose work the model was constructed. Four primary attributes of communications media that have epistemological consequences were abstracted from these works: efficiency, fixity, accessibility and transparency. The degree to which a medium exhibited high or low levels of these attributes suggested certain consequences for methods, contents, sources and validity of knowledge. This model was then applied to videotex, broadly defined as the electronic transmission of text and graphics between computer databases and remote terminals. Taken as a whole, the model indicated that videotex will tend to favor new, integrated and interdisciplinary contents, relational and heuristic methods, homeostatic storage, greater graphic notation, and a tendency to focus on contemporary, comparative data. The results of the analysis indicated that often two contrasting epistemological consequences may result from the same attribute of a medium--for example, high specialization and high interdisciplinarity of contents; high accessibility of information and high control of access. The complex nature of media change and the unformed features of the medium hindered clear projection of videotex consequences in those areas. [Note: rest will be in next issue - Ed] ------------------------------ END OF IRList Digest ********************