A01 chaf6542 06 se fm indd

  Validating a new process model 

 Whichever method has been used to arrive at the process definition, we need to check that 

the process definition is realistic. When developing a wish list of process capabilities and cor-

responding business rules the stages described by David Taylor in his book on concurrent 

engineering (Taylor, 1995) may be useful. He suggests that once new processes have been 

established they are sanity checked by performing a ‘ talk-  through,  walk-  through and  run- 

 through’. Here, the design team will describe the proposed business process as a model in 

which different business objects interact. Once the model has been adjusted, the  walk-  through 

stage involves more detail in the scenario and the design team will  role-  play the services the 

objects provide. The final  run-  through stage is a quality check in which no on- the-  spot debug-

ging occurs – just the interactions between the objects are described.    

  Data  modelling 

 Data modelling of digital business and e-commerce systems uses  well-  established techniques 

such as normalisation that are used for relational database analysis and design. As a conse-

quence, this section is brief in comparison with that on process modelling which introduces 

some novel techniques. See Bocij  et al  .  (2008,   Chapter   11   )  for  an  introduction  to  normali-

sation and relational database design. Some basic definitions are given in this section as a 

reminder of key terms. Before we start it is worth mentioning that the advent of data mining 

and  object-  oriented approaches has meant increasing use of  non-  relational database design. 

These are outlined further in  Chapters   6    and  11  of Bocij  et al  . (2008). The approach we use 

to explore data modelling for e-commerce is to use examples that identify typical elements 

of data modelling for a  sell-  side e-commerce system. We will use ER (entity relationship) 

modelling to review typical structures for these databases. In simple ER modelling there are 

three main stages.