  i* vs. AI research of intentions
In artificial intelligence research (e.g., [Cohen90][Thomas91][Lesperance91]), intentional concepts such as goals, belief, ability, and commitement have been formalized for characterizing artificial agents. However, these forumlations are not directly applicable to the i* framework, and therefore need to be adapted, for at least the following reasons [see child-branches]:(p26)
We adapt the concepts of ability and commitement as developed for agent modelling in artificial intelligence. The concepts of "routines" and "workability" are introduced to deal with the needs of strategic actor reasoning. (p26)
Unlike problem solving in AI, the objective is not to auutomatically generate some solutino, so there is no automatic chaining of means-ends. links. Each step requires the judgement and input of the modeller, as in other rationales modelling frameworks.(p41)*[Strategic Reasoning]
» See also: Strategic Reasoning
|
