153. Nag, B.N., "A Structured Model Development Framework for Knowledge-Based Systems," Dept. of Management, School of Business & Economics, Towson State University, Towson, MD, 30 pages, 12/92.
This paper explores the applicability of SM ideas to knowledge based systems.
154. Neustadter, L., "On the Structure of Data in SML Models," Research Paper, John E. Anderson Graduate School of Management, UCLA, 62 pages, 3/90.
SML stores detailed data in relational tables. Although they offer many advantages, relational tables may be subject to insertion, deletion, and update anomalies. This problem has been recognized by the database community, and the subarea of normalization theory has been developed to address it. This paper examines SML from the perspective of normalization theory. After detailing the relationship between SML and functional dependencies, conditions are given under which elemental detail tables are guaranteed to be in Boyce-Codd normal form (BCNF). It is then shown that lack of BCNF does not necessarily imply data anomalies, and a more appropriate normal form is proposed. The main conclusion is that, under mild conditions, elemental detail tables are not subject to data anomalies.
155. Neustadter, L., "Embedded SML: An Approach to Tool Integration," Informal Note, John E. Anderson Graduate School of Management, UCLA, 11 pages, 8/91.
This paper extends SML in certain ways so that it can be viewed as a "query language" for accessing information about SML models. Then, when embedded within a language like C or C++, extended SML facilitates integrating structured models with all sorts of modeling environment tools (table content rule checker, evaluator, solvers, etc.) written in or accessed through the host language.
156. Neustadter, L., "Simplifying SML: A Proposal," Informal Note, John E. Anderson Graduate School of Management, UCLA, 6 pages, 8/91. Revised 6/92 (now 11 pp.).
This note proposes changing SML's index set statement and generic calling sequence sublanguages in such a way that (1) work is transferred from the former to the latter, and (2) the resulting language overcomes certain complexities in the current design. Issues of expressive power and compatibility with SM foundations remain to be worked out.
157. Neustadter, L., "Mapping SML to SQL," Informal Note, John E. Anderson Graduate School of Management, UCLA, 10 pages, 6/92.
This note discusses how SML's default index sets can be computed by SQL queries from prior default index sets.
158. Neustadter, L., "Executable Modeling Languages as SQL-92 Applications," research proposal, School of Management, University of Texas at Dallas, 13 pages, 4/93.
This proposal, partially funded by Frontline Systems, Inc., carries on the line of investigation established in Laurel's dissertation (see Part IIA of this Bibliography). That work demonstrated how to translate a formally defined, generic, algebraic executable modeling language (EML) into a custom extension of relational algebra. Happily, the new SQL-92 standard now becoming commercially available is almost as expressive as her extended relational algebra, thereby making RDBMSs that support this standard attractive as implementation platforms for EML-based modeling environments. The first phase of this project will design, implement, and evaluate an SML to SQL-92 translator. Those few aspects that cannot be translated into SQL-92 will be handled using the C language, with the emphasis on designing a translation algorithm that leads to the most efficient expression evaluator. After making certain improvements in SML's range statement and index set statement sublanguages, the second phase will use the translator and SQL-92 to implement and evaluate a mechanism to enforce SML integrity constraints. The third phase will move on to symbolic expression evaluation and a mathematical programming interface, and the final phase will design, partially implement, and evaluate an incremental SML expression evaluator based on "derivatives" of relational algebraic expressions. [Update 9/95: the first phase uses WATCOM SQL 3.1 and is virtually finished; a paper titled "On Translating Algebraic Modeling Language Expressions into SQL" is in progress]
159. Neustadter, L., "Translating Subscripted Executable Modeling Language Expressions into an Extension of the Relational Algebra," working paper, Decision Sciences, School of Management, Univ. of Texas at Dallas, 60 pages, 12/21/94.
Based on Laurel's dissertation (listed in Part IIA of this Bibliography). She defines a generic executable modeling language called GEML that encompasses such existing subscripted languages as AMPL, GAMS, LINGO, LPL, and SML. She presents a translation algorithm from GEML to an extended relational algebra and proves it correct. An important advantage of doing this is to be able to invoke relational query evaluation and query optimization techniques that should enable more efficient expression evaluators than have heretofore been available for executable modeling languages.
160. Park, S.J., H.W. Kang and H.D. Kim, "Modeling Environment for a Manufacturing System: the Hanta Experiment," Dept. of Management Science, KAIST, Taejon, Korea, 99 pages (draft report), 5/92. Final report (in Korean), 12/92. A working paper written at KAIST based on this report is: H.D. Kim, H.W. Kang, J.W. Kim, O.B. Kwon and S.J. Park, "An Integrated Modeling Framework for Computer Integrated Manufacturing," 9/93.
This project explores the practical applicability of modeling environment concepts in the context of a production planning and scheduling system implemented in 1990 at a tire manufacturer in Korea. The system is a large one, incorporating many heuristics and policies. The experiment involves re-creating parts of the system using structured modeling and FW/SM.
The report includes the SML schema associated with this re-creation and discusses the difficulties encountered. Some of these are due to the prototype nature of FW/SM, and others are due to the nature of the model -- which in some sense is a "machine" for running the plant rather than a "model" in the traditional sense. Among the issues identified: (1) Dynamics -- extensions or supplements to SM appear to be needed to accommodate dynamic model features of the type often found in simulation modeling; (2) Top-Down Abstraction -- SM lacks this important mechanism for simplifying large models like the present one; (3) Indexing Structure Complexity -- indexing structures become quite cumbersome for this model, leading to thoughts of switching to an index-free notation.
See item 162 in this Bibliography.
161. Park, S.J. and H.D. Kim, "Constraint-Based Metaview Approach for Modeling Environment Generation," Decision Support Systems, 9:4 (June 1993), pp. 325-348.
This paper presents a paradigm to represent the structural constraints of any modeling framework together with the constraints of the functions or operations of some of the components of the modeling environment implementing the framework. The constraints are expressed in terms of objects and relationships among the objects. The description resembles a definition of predicates in first-order logic, with a set of facts and rules of inference. Indeed, a logic-based prototype system that enforces the constraints is easily generated from the predicates.
The paradigm is applied to some of the theoretical definitions of SM, such as those defining generic structure and acyclicity. In addition, examples are given to show how this approach could specify some of the functionalities (e.g., the generation of modular structure, report generation, and deletion of genera) needed for a modeling environment that supports SM.
This approach to the construction of a modeling environment appears to be flexible and able to support rich semantics of a modeling framework. A virtue of this approach is that it allows the rapid prototyping of constraints, although efficiency of execution in the generated environment appears to be a problem.
A prototype environment generator called MetaDSS has been implemented in C and INGRES on a SUN workstation.
161A. Park, S.J., J.W. Kim and H.W. Kang, "Heuristic Knowledge Representation of Production Scheduling: An Integrated Modeling Approach," Expert Systems with Applications, 10:3/4 (1996), pp. 325-339.
This paper focuses on heuristic knowledge in the context of IMF (see the dissertation by J.W. Kim and the following item on IMF in this Bibliography).
162. Park, S.J., J.W. Kim, H.D. Kim and H.W. Kang, "IMF: A Framework for Integrating Decision Models and Information Systems," Working Paper, Department of Industrial Management, KAIST, Korea, 54 pages, 1995. See also J.W. Kim and S.J. Park, "Validation Method of Mathematical Formulation in Integrated Modeling Framework," Working Paper, 11 pages, 1994. See also H.W. Kang, J.W. Kim and S.J. Park, "Integrated Modeling Framework for Manufacturing Systems: A Unified Representation of the Physical Process and Information System," International Journal of Flexible Manufacturing Systems, 10:3 (1998).
This paper presents a unified conceptual framework for both decision models and information systems, including their dynamic aspects. Although unique in many respects, the framework was influenced substantially by SM and object technology. A prototype implementation called IME has been completed using the hypertext authoring tool Toolbook, and the industrial example of Park, Kang, and Kim (item 160 of this Bibliography) is used illustratively throughout the paper.
The framework is formalized in the Ph.D. dissertation of J.W. Kim. The companion paper on validation applies to the model schema without the need for data, just as the SMLCheck and INTERP_CK processes do in FW/SM. The other companion paper elaborates IMF in the context of manufacturing systems.
163. Pollatschek, M.A., "SML for Simulation," Faculty of Industrial Engineering and Management, Technion, Haifa, Israel, 27 pages, 1995. This paper inspired another that lays the foundation for an alternative approach: "Algebraic Description of Discrete Event Simulation Models and its Implications for a Graphical Interface," 7/96. See also the thesis of his student, D. Greenberg.
This paper proposes a novel approach to dynamic simulation via structured modeling. It requires some extensions to SML -- most notably, the addition of units of measurement and random-valued attributes -- but the main idea is to begin the schema of every simulation model with a standard (the same for all models) module containing primitive and compound entities that essentially define a new simulation worldview proposed by the author. This worldview, which of course must be obeyed by the balance of the model, contains semantics beyond the representational power of SML. These semantics can be used to build a processor capable of reading any such schema and producing simulation code, in a standard target simulation language like Simscript, that gives the dynamic behavior intended by the modeler. The translation process makes extensive use of the definitional dependency structure that is inherent in SML. The resulting code can then be fed to the target language processor, which serves as the solver.
The approach is a powerful one because the schema processor can work with richer semantics than are formally expressed in the schema itself. The extra richness comes by "prior agreement" of all concerned to accept the new simulation worldview, which regards simulation as involving model entity "associations" that exist in time and that follow one another according to certain rules. This worldview seems natural, but needs to face the test of application in a wide variety of situations. The issue of generality and a pilot implementation are on the agenda for future work.
The same approach may be useful in domains other than simulation.
164. Proano, J., "Structured Modeling and Its Relevancy in Intelligent Mathematical Programming Systems," in H.J. Greenberg (ed.), Math Clinic Final Report, Dept. of Math., University of Colorado at Denver, 21 pages, Fall 1986.
Harvey Greenberg ran a "math clinic" under Electronic Data Systems sponsorship on the topic "Foundations for an Intelligent Mathematical Programming System." The final report consists of five student papers. This one sketches how EDS' PLANETS system (claimed elsewhere to have saved GM over $1 billion) can be cast within the SM framework. It also comments on how SM may contribute to the goal of achieving an intelligent math programming system.
165. Raghunathan, S., "An Application of Qualitative Reasoning to Derive Behavior from Structure of Quantitative Models," J. of Management Information Systems, 11:1 (Summer 1994), pp. 73-100.
This article uses the qualitative reasoning methodology of AI to explain certain aspects of the solution behavior of a quantitative model without having to solve it in the usual sense. This is done by studying model structure, and SML is the author's preferred structural representation. The focus is on algebraic spreadsheet and mathematical programming models. The author has implemented a prototype model processor in Common Lisp.
165A. Raghunathan, S., “A Structured Modeling Based Methodology to Design Decision Support Systems,” Decision Support Systems, 17 (1996), pp. 299-312.
The author focuses on integrating modelbases and databases as one of the main challenges of DSS design. He treats model-based DSS only (not those of a more qualitative nature). He has tested his SM-based design method in DSS classes for two years, with very satisfactory results (average project scores jumped from 60 to 85).
166. Ramirez, R., "The ASUMMS Project: An Overview," Dept. of Decision and Information Systems, Arizona State University, 10 pages, 1/20/90. Presented at the TIMS/ORSA International Meeting, Vancouver, B.C., 5/89; and at the ORSA/TIMS Meeting, Las Vegas, 5/90.
ASUMMS (Arizona State University Model Management System) is a collaborative effort involving several faculty and M.S. and Ph.D. students. The project is based on three premises: (a) SML is the sole model description language, (b) all theoretical developments must be implemented, and (c) existing software is to be used whenever possible, including relational database software (INGRES). The architecture includes user, object, relational, and physical layers. Model/solver/data independence supports the reusability of software and models. Translators map SM to SAS, IFPS, and 1-2-3. The hardware environment is a network that spans desktop computers, a VAX 8650, and supercomputers (CRAY X-MP, IBM 3090-500).
In the summer of 1991, portions of the project moved to Iowa State University with Ramirez, where it goes by the name ISUMMS.
167. Ramirez, R., "The ASUMMS Bibliography," Dept. of Decision and Information Systems, Arizona State University, 10 pages, 10/7/90.
This annotated bibliography on the ASUMMS Project at ASU comprises 3 submitted journal articles, 7 refereed proceedings, 4 technical reports, 8 informal notes, 6 Master's Applied Projects, 8 professional presentations, and 5 software documents.
168. Ramirez, R., "The SM/DB Language: Data and Manipulation Extensions for Structured Modeling," ISUMMS Project Technical Report No. 3, College of Business, Iowa State University, 31 pages, 9/91. There is also a companion paper with E. Lin, "Indexing Without Subscripts in Mathematical Programming Using Structured Modeling," Dept. of Decision and Information Systems, Arizona State University, 35 pages, 5/3/91.
This is a reference manual for the SM/DB language for structured modeling; see Lin's dissertation in Part IIA of this Bibliography for a subset now called SFL. The aim is to achieve a model management system with powerful data handling features based on a commercial relational database management system. There are sublanguages for data manipulation (including multiple model instances for a single schema) and model manipulation (including optimization), as well as model definition. The companion paper focuses on how SM/DB avoids explicit indexing.
169. Ramirez, R., "Architecture and Implementation of the DAMS System," ISUMMS Technical Report No. 4, Dept. of Management, Iowa State University, 1992.
DAMS (Decision and Algebraic Management System) is built on top of INGRES. It is accessed through sublanguages collectively called SM/DB; one based on structured modeling provides model definition and manipulation, and SQL serves as the data sublanguage. Access can be interactive or from a programming language such as C++. "Mappings" allow multiple models to share tables containing instance data. This paper is confined to a discussion of internals. [Update 9/95: the current implementation is SQL-generic, replaces INGRES with Quadbase/SQL, and stores the model catalog in the POET object-oriented database.]
170. Ramirez, R., "The ISUMMS Project and Bibliography," ISUMMS Project Informal Note No. 2, Dept. of Management, Iowa State University, 17 pages, 9/25/93.
This is an annotated bibliography covering the ISUMMS (Iowa State University Model Management System) project, a collaborative effort that carries on the prior ASUMMS project (see the previous items by Ramirez on ASUMMS).
171. Ramirez, R., "A Management System for MS/OR Models," J. of Microcomputer Applications, 1995. rtf version
This is a detailed explanation of the architecture of DAMS and its components, which aims to support large-scale modeling in a multi-user environment. A good case is made for a "data-centric" client/server architecture that emphasizes sharing existing databases with other application programs.
172. Ramirez, R., C. Ching and R.D. St. Louis, "Independence and Mappings in Model-Based Decision Support Systems," Decision Support Systems, 10:3 (October 1993), pp. 341-358. An earlier version titled "Model-Data and Model-Solver Mappings: A Basis for an Extended DSS Framework" appears in Proceedings of the First ISDSS Conference, Austin, TX, September 1990.
This paper presents the design philosophy and overall architecture of DAMS (see prior items by Ramirez in this Bibliography). DAMS makes extensive use of structured modeling for model definition, SQL for data manipulation, and the notion that model classes, instance data, and solvers should be independent of one another.
173. Ramirez, R.G. and S.E. Lin, "Subscript-Free Indexing in a Mathematical Programming Language," Proceedings of the Twenty-Sixth Annual Hawaii International Conference on System Sciences, IEEE Computer Society Press, Los Alamitos, CA, pp. 424-433, 1/93.
This paper is drawn from Lin's dissertation (see Part IIA of this Bibliography) and from Ramirez' work on SM/DB (above).
174. Ramirez, R., and R.D. St. Louis, "Model Sharing, Integration, and Multi-Purpose Application Through Data and Solver Independence," ASUMMS Technical Report No. 2, Dept. of Decision and Information Systems, Arizona State University, 45 pages, 8/6/90. Under revision.
This in-depth development of the concepts of model/data and model/solver independence uses SML for an extended discussion of the implications of these ideas for the architecture of decision support systems. The architectural conclusions were partially implemented in ASUMMS.
175. Sanchez, J., "Current Status of a Structured Requirements Methodology as Applied to Large Scale Real-Time Software Development," presented at the Workshop on Structured Modeling, UCLA, 37 pages, 8/86.
This is a progress report on a software engineering experiment at an aerospace firm aimed at developing software that can be shared among several similar radar systems. The experiment uses a structured requirements approach patterned after SM ideas. Some provocative, if tentative, conclusions are offered.
176. Sen, T., "A Graphical Approach to Model Management Using Structured Models," Pamplin College of Business, Virginia Tech, Blacksburg, VA, 27 pages, 5/90. Revised 2/92. Superseded by Chari and Sen paper listed earlier in this Bibliography.
This paper describes a limited-functionality prototype implementation of SML using HyperCard on a Macintosh. The genus graph is created using McDraw, which is resident in HyperCard. Clicking on a genus name leads to a "genus frame" that contains the same information as an SML genus paragraph. From there, one button leads back to the genus graph and another button leads to a "value frame" that contains the same information as the associated elemental detail table.
This prototype should be viewed in the context of Sen's broader research objectives, which include exploiting the superiority for some purposes of diagrammatic model representations over sentential ones. This superiority seems to arise from the adjacency of related things. In SM, calling sequences amount to adjacency information, whether or not this information is expressed diagrammatically as in genus graphs. Sen observes that such information can be used to facilitate queries about a model, and also to facilitate the implementation of some kinds of model solution procedures. This gives a new and important reason for using modeling formalisms that explicitly recognize definitional dependencies.
176A. Srinivasan, A. and D. Sundaram, "An Object Relational Approach for the Design of Decision Support Systems," Department of Management Science and Information Systems, University of Auckland, New Zealand, 1998.
Structured Modeling provides the conceptual basis for this recommended approach to DSS design, and an object relational database management system provides the preferred implementation environment. The exposition is largely example-based. See item 90 of this Bibliography, and also the next item.
176B. Srinivasan, A. and D. Sundaram, "A Visual Modelling Approach to Support Organizational Decision Making," Proceedings of the Third CAiSE/IFIP8.1 International Workshop on Evaluation of Modeling Methods in Systems Analysis and Design, Pisa, Italy, 1998.
This is a motivational discussion of the kind of DSS-building system that this research project has undertaken to build. It adopts Structured Modeling as its conceptual basis and emphasizes the importance of a visually natural user interface. See item 90 of this Bibliography, and also the immediately preceding item.
177. Stott, J.W., "Representation and Implementation of Jackson Structured Programs as Structured Models," Department of Decision Sciences, University of Hawaii, 1986.
This work was done under Army contract (TORA) in the context of combat simulation (cf. D. Dolk's 11/86 report and J.E. Kottemann's paper, both listed in this Bibliography ... note that whereas the latter paper addresses Jackson System Development, this paper addresses Jackson Structured Programming). According to the author, TORA selected SM "as an appropriate framework for model management because of its ability to describe a wide variety of models within the same environment." The purpose of the present paper is "to investigate the feasibility of embedding Jackson Structured Programming (JSP) representations within Structured Modeling representations."
Stott asserts that JSP is built on the idea that a computer program is a model of reality that should resemble its target reality as much as possible. This idea, which arises also in connection with the object-oriented programming paradigm, is one of the connections by which programming languages and modeling languages can cross-fertilize one another.
Stott explores three different approaches for representing JSP's using SM. None are very successful. The execution of the three approaches leaves much to be desired in terms of SM technique, but that is to be expected given the lack of decent tutorial materials on SM at the time this paper was written (apparently, even the working paper version of my 1987 Management Science article was not available to the author). In spite of this handicap, the final two sections of the paper are insightful.
178. Tsai, Y., "An Operational Approach to Model Integration Using a Structured Modeling Framework," Proceedings of the 1993 Pan Pacific Conference on Information Systems, Kaohsiung, Taiwan. Full version: Research Paper, John E. Anderson Graduate School of Management, UCLA, 60 pages, 12/87; revised 2/88. Derivative paper "Model Integration Using SML" published in Decision Support Systems, 22:4 (April 1998).
This paper defines several operations on an SML schema, or on a pair of SML schemata, that are of particular interest for formalizing model integration. These begin with low level operations on the tree of paragraphs composing a schema and progress through higher level operations. The most interesting operations are schema "projection" (to extract the minimal subschema needed to support a given subset of genera), the "joining" of two schemata (to marry schemata having certain pairwise commonalities), and three different ways to revise the modular structure (or its ordering) of a schema in an attempt to achieve monotonicity.
Executing any of these operations could possibly result in a schema that violates the Schema Properties (static semantics) of SML as set forth in my WMSI Working Paper 360. The paper studies this issue in detail for four properties: (a) no /ce/ or /a/ genus may call an /a/, /f/, or /t/ genus; (b) genus names must all be distinct; (c) each module must include at least one genus paragraph; and (d) every called genus must come before the calling genus in the schema. Seven main propositions and several examples settle virtually all of the open questions along these lines for the operations studied.
This paper contributes to the important task of understanding how different kinds of schema edits can disrupt the static semantics of SML, and what can be done about it. It helps lay the foundation for the eventual development of an incremental static semantic analyzer and smart schema editor.
179. Tung, L., "Relational Representations of Structured Modeling," Dept. of Decision and Information Systems, Arizona State University, 20 pages, 2/1/90. Earlier version presented at the ORSA/TIMS International Meeting, Vancouver, B.C., 5/89.
This paper, based on a Master's Applied Project, proposes a particular way to store an SML schema in the form of relational tables, a design problem that others have also addressed (including Dolk, Lenard, and Tsai). This particular design was adopted by ASUMMS and then by ISUMMS (see Ramirez's items on these systems earlier in this Bibliography). The design is evaluated according to functionality and efficiency criteria, especially by comparison with Dolk's IRDS proposal.
180. Tung, L., R. Ramirez and R.D. St. Louis, "Model Integration in an Object-Oriented Environment," Proceedings of the Twenty-Fourth Annual Hawaii International Conference on System Sciences, Vol. III, IEEE Computer Society Press, Los Alamitos, CA, pp. 284-290, 1/91.
This paper extends the development of item 8 of this Bibliography.
181. Worobetz, N.D. and G.P. Wright, "ORACLE Structured Modeling (OR/SM): User's Manual," and "ORACLE Structured Modeling (OR/SM): System Manager's Manual," Version 1.1, Krannert Graduate School of Management, Purdue University, respectively 44 pages and 54 pages, 10/22/90.
OR/SM (ORACLE/Structured Modeling) is implemented in about 60,000 lines of C using LEX and YACC, with ORACLE as the host system. It aims to support model management, statistical analysis, and model integration. Models are represented in a close dialect of SML, and the user interface is completely menu-driven. Statistical analysis of the data, also menu-driven, is carried out via a largely transparent interface with PC-SAS. Users can make data queries using the full power of SQL.
The first document details the menu interface of OR/SM from the point of view of the end user. The second document does the same thing from the point of view of the system manager. For further details, see the item by Wright, Worobetz, Kang, Mookerjee and Chandrasekharan later in this Bibliography.
This and related work was funded by two grants: (a) AT&T Foundation grant for "Modeling and Analysis Using Structured Modeling," 1989-91, and (b) NATO Collaborative Research Grant (with Peter Mevert, Free University of Berlin) for "Network Modeling Using Structured Modeling," 1989-92.
182. Wright, G., "An Integrated Marketing Information System Based on Structured Modeling," slides from a presentation at the Workshop on Structured Modeling, UCLA, 15 pages, 8/86. Progress report presented in the session on SM at the TIMS/ORSA Meeting in New Orleans, 5/87.
Imagine a modeling environment in which you have, in proper SM form, these things insofar as they pertain to your work: (a) the main commercial databases, (b) the most important data for your organization, and (c) the classical models of your discipline. Then when you want to do some analysis or a model-based application you may be able to customize a classical model or two and obtain much of the necessary data from your handy databases. This is the sort of environment to which Gordy aspires for marketing applications. Similar systems would be worth building in many other functional areas (finance, logistics, production, scheduling, etc.).
183. Wright, G., N.D. Worobetz, M. Kang, R. Mookerjee and R. Chandrasekharan, "OR/SM: A Prototype Integrated Modeling Environment Based on Structured Modeling," INFORMS J. on Computing, 9:2 (Spring 1997), pp. 134-153. Also available as a working paper, Krannert Graduate School of Management, Purdue University, 36 pages, 8/94. See also the companion paper by Kang et al. earlier in this Bibliography on the design and implementation of OR/SM. And see part 3 of R.V. Mookerjee, Three Essays on Model Management, Ph.D. Dissertation, Purdue University, 116 pages, 1993. Supervisor: G. Wright.
This paper motivates and describes the functionality of OR/SM (ORACLE/Structured Modeling), a remarkably ambitious system catering to statistical databases and operations management applications. The subject of Worobetz' doctoral dissertation (see Part IIA of this Bibliography), OR/SM is built on an industrial-strength platform: ORACLE Tools and Database. Models are composed in Microsoft Word using SML extended to include matrix-valued generic rules, with parts of level 4 SML omitted and with minor modifications to exploit SQL's potential for efficient evaluation and retrieval. Expression evaluation is done largely through SQL. The schema and data are input through user interaction within a completely menu-driven system. Statistical data analysis, also menu-driven, is carried out via a largely transparent PC-SAS interface, and a SAS/OR interface handles optimization models. An interface to the commercial package QS facilitates solving a variety of operations management problems. OR/SM is extensively documented, and has been used successfully by many students.
183A. Yang, De-Li and Xiatong Wang, “An Approach to Unify Model Definition and Manipulation,” Computers and Industrial Engineering, 33:3,4 (Dec. 1997), pp. 661-663.
This paper proposes a hierarchical spreadsheet representation for SM that seeks to integrate model representation, model manipulation, and database management. Moreover, it is compatible with functional programming and relational database query languages.
184. Yeo, Gee Kin and Guoqiang Li, "On Discrete Time Modeling," Proceedings of IASTED International Conference on Modelling, Simulation and Optimization, Gold Coast, Australia, May 6-9, 1996. A 32 KB Word 6.0 file is available by ftp here. Expanded subsequently as Technical Report TR31/96, Dept. of Information Systems and Computer Science, The National University of Singapore, 21 pages, 11/96. A 213 KB Word 6.0 file is available by ftp here.
This paper examines the role of time in MS/OR modeling, and suggests adding a new genus type to SML that, unlike the original genus types, has a fixed semantic interpretation relating to time. The second author is writing a graduate thesis on this topic.
184A. Yeo, Gee Kin and Jian Hu, “Visual Modeling with VMS/SM,” Proceedings of IASTED International Conference on Simulation and Modelling, Pittsburgh, pp. 202-205, 5/97. A 575 KB Word 6.0 document is available by ftp here. For earlier papers on VMS/SM, see Hu and Yeo, "Visualization of Structured Modeling," Proceedings of the Second Pan Pacific Conference on Information Systems, pp. 342-348, 7/ 95 (a 53 KB Word 6.0 file is available by ftp here) and also Yeo and Hu, “VMS/SM, A Two-tiered Visual Modeling System,” Technical Report TR21/96, Department of Information Systems & Computer Science, The National University of Singapore, 22 pages, 11/ 96 (a 80 KB zipped Word 6.0 file is available by ftp here).
The authors have supervised the implementation of an ambitious prototype visual modeling environment called VMS/SM (for Visual Modeling System for Structured Modeling). This paper is a brief introduction to VMS/SM, which implements a major subset of structured modeling. The other two papers cited give more design details.
See item 62A of this Bibliography for additional detail and here for extensive information concerning this system, including a tutorial.
185. Yusof, M., T.M.T. Sembok and M.M. Rahim, "An Approach to the Development of a Decision Support System Tool," Dept. of Computer Science, Universiti Kebangsaan Malaysia, Selangor, Malaysia, 20 pages, 3/92. Presented at the XI Southeast Asia Regional Computer Conference (Kuala Lumpur, Malaysia, 11-14 August 1992).
This paper proposes a straightforward way to use structured modeling as an approach to the early stages of model formulation. It also suggests some model complexity indices that can be computed from a genus graph.