In order to ensure that an information system gives the bestpossible support to a business, it is necessary to analyse thebusiness. This is required, not primarily for understanding thebusiness as such, but in order to enable stakeholders tofortnulate precise requirements which, with reasonablecertainty, wIll result in a supportive information system. Inorder to facilitate acquisition of requirements, knowledgeabout the business as well as about the requirements needs tobe represented in a precise and understandable way. This allowsstakeholders and analysts to communicate effectively.
In this thesis a set of languages that support modelling ofbusinesses as well as of requirements of information systems isdefined. The languages are each specialised on one type ofknowledge of the business, e.g., knowledge about goals,concepts, behaviour, and rules. The graphical representation ofthe languages is intended to be easy to understand at theexpense of expressive power. Still, the languages need to bepi~cisely defined in order to reduce risk for unintentionalambiguities. It is assumed that the development process can bedivided Into three phases (or worlds) Subject World Modelling(SWM), Information System World Modelling (ISWM) andImpementation World Modelling (IWM).
Therefore, each language is defined by a mets model thatdefines its syntactical properties. Several formal meta rulesare presented for these languages and their inter languagerelationships. This makes It possible to represent new kinds ofknowledge in the specification that will reduce the number ofpermissible interpretations of a specification. In this way thesemantic quality of specifications can be increased as well.Finally, this thesis also discusses a number of methodologicalaspects of the languages presented.
Kista: Data- och systemvetenskap , 1997. , 338 p.