5.1 Introduction

Before introducing the selected approaches, some reflections are made on the criteria used for selecting and evaluating them.

5.1.1 Criteria for selection and evaluation

The criteria used for selection and evaluation of approaches to enterprise modeling are first and foremost related to the research questions posed in section 1.2.2. In addition, a certain variety in background (e.g., commercial versus academic) is sought.

First criterion: Understanding is the primary purpose

The first criterion used for selection is purpose of enterprise modeling. As stated in the main research question, the focus in this thesis is enterprise modeling for human sense-making and communication. Hence, the selected modeling approaches must have as a primary purpose to make people understand more of the domain.

However, few existing approaches have this purpose as their primary one and consequently, approaches are discussed even if they list other purposes as well. A problem with these may be that the more diverse purposes each framework have, the more diverse requirements they will have to fulfill. Requirements may be contradictory in the sense that fulfilling one of them inhibits the fulfillment of others, i.e., they are interdependent. The first criterion concerns RQ1.

Requiring executability of models require that some aspects of the modeling language are formally defined. Formality may reduce the ease of initial model formulation through demanding strict adherence to consistency rules.

Second criterion: Focus on both models and modeling

A framework may focus on properties of the final model and/or properties of the modeling process. From the research questions of section 1.2.2 (particularly RQ2 and RQ3), focus on both the modeling process and on models is required. Seeing modeling as social construction of reality is not compatible with neglecting the modeling process.

Third criterion: Type of work that is modeled

The taxonomy of different types of work can be used to discuss the appropriateness of various frameworks. A modeling approach developed particularly for one type of work might be less suited for modeling of other types of work. Type of work is related to the areas discussed in chapter 4, and is concerned by RQ4.
An enterprise modeling approach suitable for the UoD manufacturing of cars may have an ontology that describe automobile parts in detail, e.g., auto body, paint qualities and motor parts. Trying to apply the approach to modeling of office work might at best leave a large part of the ontology superfluous, and at worst not provide necessary expressiveness for phenomena frequently observed in an office, e. g., business letters and insurance claims.

5.1.2 Frameworks included

The following nine frameworks are presented and discussed: The frameworks comprise a mix of academic research (F3, TOVE, SSM and SAMPO), industrial projects (Enterprise and Caesar) and commercially available approaches (Metis, OMT and OOram). The last four in the list are not dedicated enterprise modeling approaches, but exhibit particular qualities defending their inclusion. To anticipate the discussion, F3, SSM and OOram are found to be the most inspiring approaches and consequently, they are discussed in some more detail than the others.

5.1.3 Presentation of the frameworks

The frameworks are discussed in terms of the following list of features: A brief evaluation and summary of all modeling approaches are provided in the closing of the chapter.
A note on the assessment: Most of the properties can be readily assessed as the authors of the frameworks explicitly state their position on purposes, methods, models, etc. Other features, like type of work and in particular worldview, are seldom explicitly discussed in the surveyed literature. 

This raises the question of how the lack of something is to be interpreted and assessed. The stance taken here is that lack of an explicit position on an issue is an indication that the issue is considered insignificant. The presentation below provides whatever evidence found and at the same time admits that the conclusions drawn at times are uncertain. The intention of making the assessments is not to categorize approaches, but rather to discuss the approach in light of the category. The explicit assessment is made to summarize the discussion.