Decision modeling is a successful technique that develops a richer, more complete business understanding earlier. Using decision modeling for analytic project requirements enables organizations to:
- Compare multiple projects for prioritization, including allowing new analytic development to be compared with updating or refining existing analytics.
- Act on a specific plan to guide analytic development that is accessible to business, IT and analytic teams alike.
- Reuse knowledge from project to project by creating an increasingly detailed and accurate view of decision-making and the role of analytics.
In DecisionsFirst Modeler, a predictive analytics or data mining result is captured as a Knowledge Source (KS). The link between a decision and a KS are authority requirements (dashed line).
A first level of refinement, you can identify authority requirements for a KS, for example the data you would look at to derive the predictive analytics or data mining result. For some analytic projects where more detail about the actual analytic project is needed, you would create an analytic effort object and link it to the relevant KS.
Here is a short video on creating an analytic effort object .
Also in this webcast, starting at about 24', James describes decision requirements. Here you will see the high level decision with some analytic KS, and then the decision decomposition, etc.