Font Size: a A A

Subscript free modeling languages: An improved basis for integrating models and data in decision support systems

Posted on:1994-11-03Degree:Ph.DType:Dissertation
University:Arizona State UniversityCandidate:Lin, Suh-Yun ElvaFull Text:PDF
GTID:1478390014494176Subject:Operations Research
Abstract/Summary:
Model management systems (MMSs) are becoming more important to decision makers (DMs). Two reasons for this increased importance are: (1) decision models now are recognized as organizational assets, and (2) the problems faced by DMs are so complex they can be solved only with models. Modeling languages have been introduced to facilitate model use and development. Most modeling languages use subscripts. However, there are some problems with using subscripts. First, subscripts make an expression more complicated and harder to understand than non-subscripts. Second, the use of subscripts increases the difficulty of representing problems with expressions. Third, subscripts reduce the speed of processing an expression.; A new modeling language SM/DB (Structured Modeling/DataBase) is designed in this research to eliminate these problems. One unique feature of SM/DB is that it does not use subscripts. Some advantages of subscript-free languages are: (1) readability is enhanced; (2) DMs can more easily express a problem, and (3) parallel processing is facilitated.; Another unique feature of SM/DB is the linkage with a Relational Data Base Management Systems (RDBMSs). This linkage not only can integrate a MMS with a RDBMS but also allows a MMS to be used with a RDBMS. That is, SM/DB models can be stored in a RDBMS and these models draw inputs from a RDBMS and return outputs to the RDBMS. The use of a RDBMS provides an efficient way to store and to manipulate data and models. In SM/DB, the linkage is done by translating the SM/DB statements into Structured Query Language (SQL).; The expressiveness of SM/DB is evaluated in three ways. First, SM/DB is shown to be able to represent all of the major types of index sets and indices. Second, it is shown that SM/DB supports Structured Modeling (SM). Third, the expressive power of SM/DB is shown to exceed the expressive power of Geoffrion's Structured Modeling Language (SML), which uses subscripts.; Overall, the conclusion is that SM/DB is more expressive and more user friendly. It provides far more effective and efficient model management than existing subscript modeling languages.
Keywords/Search Tags:Modeling languages, SM/DB, Decision, Models, RDBMS, Management, Data
Related items