Data flows: Note on Data-Driven Process Modeling

A Focus on Data
Data flow diagrams (DFDs)
Common DFD mistakes
Summary/next steps


  1. In the least attractive example of such a redesign, "automation" is used to make a process more efficient ("Your orders handled in 1 day!") while greatly reducing the number of people needed to complete the process (e.g., three-quarters of the non-management staff in the organization are fired). The results of projects such as this have raised a debate between academics and other assorted pundits about the social utility of computing (indeed, several prominent academics built their careers on the social implications of automation long before "ethics" became a recognized part of business school curricula). We won't go into this debate here except to note that the emphasis in this course is placed very strongly on systems that enhance the expertise of people who currently work within an organization rather than on automation systems that replace variable-cost labor with fixed-cost capital. The effect is the same - do a greater volume of business using in-house resources - but the means are different (an enhancement approach does not threaten to strip an organization of existing and potentially valuable human expertise).
  2. In this representation, the dashed box in the upper lefthand corner of the Source/Sink symbol refers to an identification number. In Computer-Aided Systems Engineering software packages (CASE), it is possible to derive database specifications and programming code directly from suitably detailed DFDs and ERDs. In such packages, each DFD symbol receives its own id number.
  3. Black hole is a highly technical term which has been developed through years of research. Some evidence exists, in fact, to support the hypothesis that the existence of black holes in distant galaxies implies that God screwed up several early DFDs.