Authors:
Nelufar Ulfat-Bunyadi
;
Nazila Gol Mohammadi
and
Maritta Heisel
Affiliation:
University of Duisburg-Essen, Duisburg and Germany
Keyword(s):
Domain Knowledge, Assumption, Domain Hypothesis, Context, Context Modelling.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Knowledge Management and Information Sharing
;
Knowledge-Based Systems
;
Requirements Engineering
;
Symbolic Systems
Abstract:
In requirements engineering, different types of modelling techniques exist for documenting requirements and their refinement (e.g. goal-oriented techniques, problem-based techniques). Each type of technique has its advantages and shortcomings. However, extensions made to one type may be beneficial to another type as well, if transferred to it. KAOS is, for example, a comprehensive methodology that supports goal-oriented requirements engineering. As part of the KAOS methodology, multi-agent goals are refined until they can be assigned to single agents in the software or in the environment. Beside goals, domain properties and hypotheses (facts and assumptions about the environment) can also be modelled in KAOS goal models as well as their influence on the satisfaction of goals. However, the KAOS methodology provides limited support in the systematic refinement of goals. Developers using the KAOS method are left alone in refining the multi-agent goals and in making domain properties and
hypotheses explicit. The Six-Variable Model, on the other hand, is an extension of problem diagrams and supports a systematic refinement of requirements and a systematic elicitation of domain properties and domain hypotheses. In this paper, we show how the Six-Variable Model can be used to support a systematic refinement of goals in KAOS goal models.
(More)