IRList Digest Friday, 26 August 1988 Volume 4 : Issue 44 Today's Topics: Email - Changes in addresses for IRList news Call for Papers - ACL 1989 Annual Meeting; Vancouver, 26-29 June Abstracts - Dissertations selected by S. Humphrey [Part 1 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: Fri, 26 Aug 88 18:19:22 EDT From: Edward A. Fox Subject: email changes 1) The BITNET node I use, vtcc1, should now be listed officially in all host tables. Thus, people should be able to reply to me and to send there. Please stop using "foxea@vtvax3" unless you have problems, in which case our systems people would like to know and should be able to correct the situation. 2) If your mail goes through some system understanding name servers you should be able to mail directly to me at: fox@fox.cs.vt.edu If you cannot, you definitely can reach me at fox%fox.cs.vt.edu@dcssvx.cc.vt.edu Please be aware that vtopus.cs.vt.edu will be down starting next Tuesday for probably a week or two, or maybe even longer, so please switch your mail to one of these addresses. Thanks, Ed ------------------------------ Date: Mon, 1 Aug 88 10:30:14 EDT From: Donald E Walker Subject: ACL 1989 Annual Meeting Call for Papers; Vancouver, 26-29 June CALL FOR PAPERS 27th Annual Meeting of the Association for Computational Linguistics 26-29 June 1989 University of British Columbia Vancouver, British Columbia, Canada TOPICS OF INTEREST: Papers are invited on substantial, original, and unpublished research on all aspects of computational linguistics, including, but not limited to, pragmatics, discourse, semantics, syntax, and the lexicon; phonetics, phonology, and morphology; interpreting and generating spoken and written language; linguistic, mathematical, and psychological models of language; machine translation and translation aids; natural language interfaces; message understanding systems; and theoretical and applications papers of every kind. REQUIREMENTS: Papers should describe unique work that has not been submitted elsewhere; they should emphasize completed work rather than intended work; and they should indicate clearly the state of completion of the reported results. FORMAT FOR SUBMISSION: Authors should submit twelve copies of an extended abstract not to exceed eight double-spaced pages (exclusive of references) in a font no smaller than 10 point (elite). The title page should include the title, the name(s) of the author(s), complete addresses, a short (5 line) summary, and a specification of the topic area. Submissions that do not conform to this format will not be reviewed. Send to: Julia Hirschberg ACL89 Program Chair AT&T Bell Laboratories, 2D-450 600 Mountain Avenue Murray Hill, NJ 07974, USA (201)582-7496; julia@btl.att.com SCHEDULE: Papers are due by 6 January 1989. Authors will be notified of acceptance by February 20. Camera-ready copies of final papers prepared in a double-column format, either on model paper or in a reduced font size using laserprinter output, must be received by 20 April along with a signed copyright release statement. OTHER ACTIVITIES: The meeting will include a program of tutorials organized by Martha Pollack, AI Center, SRI International, 333 Ravenswood Avenue, Menlo Park, CA 94025, USA; (415)859-2037; pollack@ai.sri.com. Anyone wishing to arrange an exhibit or present a demonstration should send a brief description together with a specification of physical requirements (space, power, telephone connections, tables, etc.) to Richard Rosenberg at the address below. CONFERENCE INFORMATION: Local arrangements are being handled by Richard Rosenberg, Department of Computer Science, University of British Columbia, Vancouver, BC, CANADA V6T 1W5; (604)228-4142; rosen%cs.ubc.ca@relay.cs.net. For other information on the conference and on the ACL more generally, contact Don Walker (ACL), Bellcore, MRE 2A379, 445 South Street, Box 1910, Morristown, NJ 07960-1910, USA; (201)829-4312; walker@flash.bellcore.com or bellcore!walker. PROGRAM COMMITTEE: Joyce Friedman, Barbara Grosz, Julia Hirschberg, Bob Kasper, Richard Kittredge, Beth Levin, Steve Lytinen, Len Schubert, Martha Palmer, Fernando Pereira, Carl Pollard, Mark Steedman. ------------------------------ Date: Wed, 3 Aug 88 13:36:58 EDT From: "Susanne M. HUMPHREY" Subject: dissertation abstracts Ed, latest edition of dissertations for irlist and SIGIR Forum. Please forward to Vijay for me. I left in the .[]. delimiters in case they were useful to you; let me know for future. [Note: I have left them in too - Ed] ... --Susanne .[ AN University Microfilms Order Number ADG87-28544. AU DILLA, WILLIAM NOEL. IN The University of Texas at Austin Ph.D 1987, 206 pages. TI TESTS OF INFORMATION EVALUATION BEHAVIOR IN A COMPETITIVE ENVIRONMENT. DE Business Administration, Accounting. AB This study addresses the question: "How does context affect the behavior of information system evaluators in a competitive environment?" using an experimental setting. Four different explanations of why context might or might not affect this behavior are proposed and examined. They include the expected utility hypothesis and three explanations based on recent research on cognitive processes in judgment and choice: (1) individuals act according to prospect theory, (2) individuals act with information because they wish to avoid uncertainty, and (3) individuals view information as a valuable good. This research extends previous work on subjective evaluation of information systems in two ways. First, earlier studies either were strictly descriptive or sought to build simplified combinatorial models of subject behavior. The present study goes beyond this to examine the decision strategies individuals employ for subjective information evaluation. Second, previous work examined settings with no conflict of interest between individuals, using a decision-theoretic framework. The present study examines a competitive two person setting, using a game-theoretic framework. This scenario is more representative of environments in which accounting information is actually used than the settings of earlier experimental work. In the experiment, subjects played the part of one of two managers in a firm operating in a simplified production environment. They interacted with a microcomputer, which played the part of a second manager. The subjects had the option to obtain private information on a random variable affecting the production process. Contextual variables manipulated were task setting and stated cost of information. The principal findings of the study are that individuals: (1) apparently use simplified decision strategies in the information evaluation problem and (2) tend to ignore the effects their private information will have on the actions of the other manager. A significant proportion of subjects consistently obtained information in cases where it had negative expected value. This proportion was unaffected by task setting, indicating individuals choose to act with information because they wish to avoid uncertainty. Some subjects' choices were also affected by stated information cost. This effect differed across task settings. .] .[ AN University Microfilms Order Number ADG88-01444. AU WESTLAND, JAMES CHRISTOPHER. IN The University of Michigan Ph.D 1987, 259 pages. TI SEMANTIC NETWORKS: A STOCHASTIC MODEL OF THEIR PERFORMANCE IN INFORMATION RETRIEVAL. DE Business Administration, Accounting. AB Recent advances in computer technology have made possible the implementation of information retrieval strategies adopting semantic network approaches from artificial intelligence. This research investigates the claim that document based information retrieval strategies which automatically extend the user's query by inferring, through the use of a semantic network, the concepts that the user wishes to retrieve, can outperform traditional Boolean retrieval strategies. The claim reflects a growing concern among researchers and commercial information retrieval services that many relevant documents are not being retrieved in systems using Boolean retrieval strategies. Research by Furnas, and Blair and Maron attribute this to problems in user query formulation, particularly the use of multiple terms to express similar meanings. This research compares two classes of document based information retrieval strategies--Boolean and "concept" extended. The latter strategy is adapted from research by Kochen on semantic network based document retrieval. The Kochen algorithm generates an equivalence class of terms called "concepts" from a semantic network which are used to disjunctively extend a document database query. This research analyzes Boolean and "concept" extended retrieval strategy performance via a stochastic model of user querying and information retrieval. Stochastic processes for querying, retrieving and constructing semantic networks are proposed, and resulting probability distributions are constructed in the analysis. The research also investigates various alternatives for measuring the performance of document retrieval systems. The research results show improved performance measured by most performance statistics for the "concept" extended retrieval strategy over the traditional Boolean strategy. "Concept" extended queries were shown to improve expected recall as much as 20%, and expected precision as much as 10%. This is attributed to that strategy's injection of new information about document database content into the user's query. The research also provides evidence for the positive correlation of precision and recall, and for the superior informativeness of precision and recall over composite and Neyman-Pearson measures of retrieval performance. The research additionally provides a limited microeconomic analysis of the costs and benefits of document based information retrieval strategies in a competitive marketplace. This provides economic guidelines for commercial implementations of document retrieval systems using "concept" extended retrieval strategies. .] .[ AN University Microfilms Order Number ADG87-19970. AU HAREL, ELIE C. IN University of California, Los Angeles Ph.Do 1987, 219 pages. TI IREX: AN AUTOMATED INFORMATION REQUIREMENTS EXPERT. DE Business Administration, Management. AB This dissertation addresses the problem of information requirements generation as part of the business strategic planning process. Currently, such processes require elaborate manual analysis that needs to be performed by highly skilled and experienced consultants. The current procedures, therefore, are extremely lengthy and expensive. Furthermore, the documented knowledge about these methodologies is deficient and incomplete. The methodology proposed in this dissertation takes the approach that the information requirements generation process can be made simpler, faster, and less expensive, by its consolidation into an expert system that would be used as a tool by information system analysts and possibly end users. Such a system will incorporate significant parts of the knowledge accumulated by acknowledged experts in this field, as well as industry specific information, and a sophisticated easy-to-use human interface. In addition, new knowledge about the information requirements generation process can be obtained and documented by the knowledge acquisition phase of the expert system creation. IREX (Information Requirements EXpert) is a prototype expert system for generating business information requirements. It uses a variation of the BICS methodology which assumes the existence of a finite set of generic models which can be customized for all businesses. IREX incorporates knowledge acquired from experts who have been using BICS over an extended period of time, and is currently tailored for the use by real estate brokerage firms. IREX was implemented and tested on a microcomputer using the PROLOG language. The tests were conducted in a real world environment where specific information requirements were generated for several real estate brokerage firms. The results show that the concepts behind IREX lend themselves very nicely to the expert system approach. The problems encountered were mostly due to semantics to which the users were not accustomed, and help facilities that were not fully implemented within the prototype version of IREX. The results also indicate that the BICS methodology is not as naturally applicable to service oriented enterprises as it is to production or manufacturing type firms. Continued research is being done to enhance IREX as practical tool for information systems planners in multiple industry types. (Abstract shortened with permission of author.). .] .[ AN University Microfilms Order Number ADG87-28739. AU LE, LAN THI. IN University of Houston Ph.D 1987, 361 pages. TI DEVELOPMENT OF AN EXPERT SYSTEM FOR GAS CHROMATOGRAPHY - MASS SPECTROMETRY HYDROCARBON DATA INTERPRETATION AND AN EQUILIBRIUM MODEL FOR THE EFFECT OF NEUTRAL SURFACTANTS IN REVERSED-PHASE HIGH PERFORMANCE LIQUID CHROMATOGRAPHY. DE Chemistry, Analytical. AB An expert system computer program has been developed to automatically arrive at decisions normally made by a highly specialized highly skilled analytical chemist in the routine interpretation of data from a Gas Chromatography/Mass Spectrometry (GC/MS) data system. The specific purpose of the expert system is to set chemical limits within which the system accepts one or more of nine candidate compounds ("spectral matches") provided by a commercially available data system. This is accomplished by assuming that each gas chromatographic peak might be a mixture; deconvoluting each peak using mass spectral "type analysis" into as many as 3 possible components of each of as many as 8 different types; and assigning a carbon number range to each possible component based on its elution time. A library identification is assigned if and only if (a) it is within these type and carbon number limits, and (b) the corresponding deconvoluted spectrum also show a good "match" or "fit" with the library spectrum. If the library identifications are unacceptable, the component is assigned a generic identification in terms of component type and carbon number. The expert system was constructed to retain the human expert's knowledge and to reduce the need for the expert whose specialized skills can be used more productively in non-routine areas. The effect of solute ionization and the concentrations of nonionic and zwitterionic (neutral but charged) surfactants on the retention times of weak acids, bases, and ampholytes in reversed-phase high performance liquid chromatography was investigated. An equilibrium model based upon interaction among eluent and solutes is derived to predict the retention behavior. Micro-Bondapak C$\sb{18}$ was used as the stationary phase, and an aqueous buffer solution was used as the mobile phase in which nonionic and zwitterionic surfactants were added at different concentrations for each pH. Retention times of solutes were found to decrease as the concentrations of nonionic surfactants in the mobile were increased. The same results were obtained from the weak acids and ampholytes in zwitterionic surfactant, but not for weak bases. At high pH, the retention times of weak bases were decreased as the concentrations of zwitterionic surfactant were increased. At low pH, the retention times of weak bases were found to increase as the concentrations of zwitterionic surfactant were increased. It is hypothesized that zwitterionic surfactants are oriented on the surface in such a way that non-equivalent charge-charge interactions occur with charged samples. .] [Note: continued in next issue - Ed] ------------------------------ END OF IRList Digest ********************