5.2.3 Toronto Virtual Enterprise (TOVE)

The TOVE project at Enterprise Integration Lab, University of Toronto, develop a framework for enterprise integration (EI) based on enterprise modeling. The main focus of EI is on improved communication and coordination within and between organizations in order to achieve higher levels of productivity, flexibility and quality (Fox et al., 1993).

Enterprise models developed according to the TOVE framework are meant to (ibid.):

  1. provide a shared terminology for organizational agents to understand and use,
  2. define meaning of each term in a precise and unambiguous manner,
  3. be a basis for automated deduction of "common sense" questions about an enterprise, and
  4. define a symbology for graphical visualization of terms.
In their discussion of enterprise models, Fox et al. (1995) take what they refer to as a "second generation knowledge engineering approach" to the construction of enterprise models. A first generation KE approach is extracting rules from experts, while second generation is ontology engineering: They develop comprehensive ontologies for all the aspects of an organization they find necessary (necessity is decided based on competency requirements to the model, i.e., what are the questions the model will have to answer, either by ordinary look-up or by deduction). The background of TOVE is clearly KE and to some degree CIM.

Papers reporting on developments and application of TOVE exclusively focus on characteristics of enterprise models, i.e., ontology elements and their representation using first-order logic. Enterprise modeling as a process is not considered to any significant degree, and a reasonable conclusion is that it is not considered problematic.

There is significant evidence that the underlying ontological and epistemological foundation of TOVE is mainly objectivistic. Firstly, there is the lack of discussion of enterprise modeling process. An example is the claim made by Fox et al. (1993) that

"to achieve integration it is necessary that units of the enterprise, be they human or machine based, be able to understand each other. Therefore the requirement exists for a language in which enterprise knowledge can be expressed."
They recognize the need for human understanding, but neglect the process of developing it. A language for developing models is the most important. This is a tradition that Bubenko (1993:73) claims to be characteristic of traditional methods and techniques for information modeling.

Secondly, their strong focus on ontologies of the domain of enterprise models indicates that they assume general validity and relevance of their concepts, indifferent of the reality construction process of the modeling organization.

Finally, the degree of formality and the notation used for modeling is accessible to specialists only, and does not lend itself to intuitive use by non-experts in first-order logic. Hence, the language they provide is not meant for direct use in the sense-making part of the modeling process, it must be used by specialists (restricting access to the construction process).