AN INITIAL READING PATH FOR STRUCTURED MODELING

Arthur Geoffrion
UCLA Anderson School of Management

 
When people ask me what they should read to obtain a decent understanding of Structured Modeling (SM), my response usually is to suggest a reading path like the following. It aims to economize on reading time.  But if they just want to read a couple of pages, I send them to this encyclopedia article.
  1. "Introduction to Structured Modeling" (1987 Management Science article)
    2. This overview of structured modeling ideas is a suitable starting point. Read the whole thing without much attention to details. Since writing this article, I have (a) discovered that a level-by-level explanation of computationally active definitional dependency diagrams is the best way to explain SM's "universal" modeling paradigm, (b) formalized the SML language, (c) completed the FW/SM research prototype, and (d) done much related research.
  2. "Formal Aspects of SM" (1989 Operations Research article)
    3. This paper formalizes the SM modeling paradigm. Read Section 1, and then skim Sections 2 and 3 together with the figures in the appendix.
  3. "The SML Language for SM" (Operations Research articles)
    4. This pair of papers provides an example-based description of a notational system for SM, together with pointed discussion of SML's allegedly "notable characteristics". Read both papers straight through without much attention to language details. Keep in mind that SML is just one of many possible notational systems for the SM modeling paradigm; for example, I would like to see a hybrid notational system that is more explicitly graphical and hence closer to the underlying modeling paradigm.

    To learn more about SML, the next step would be my Working Paper 378, which is a long tutorial on the subject. Copies are still available.

  4. "FW/SM: A Prototype SM Environment" (1991 Management Science article)
    5. This paper describes the functionality of an SML implementation built on top of Framework, and discusses the significance of certain features of general interest. There is also a more technical companion paper by Neustadter et al. giving details about design and implementation.
  5. "Computer-Based Modeling Environments" (1989 EJOR paper)
    6. This can be read independently of the above papers. Although it is not on Structured Modeling per se, it details much of my philosophy about the sort of modeling environment that will some day transform model-based work. Some of the references to then-current work are now a bit dated, but I still believe as strongly as ever in the views expressed in this paper.
At this point the reader will have a pretty good understanding of SM, and may be ready to read further in the SM literature or to look for open research topics. See: