Revisiting quality metrics for conceptual models

Determining quality of a conceptual model has always been a challenge for researchers. Various quality attributes and their measures have been proposed in the research. These measures are usually qualitative and lack quantitative aspect which makes it difficult to discuss quality objectively. In our research, we have studied the existing quantitative metrics and have found that many of these metrics are not appropriately defined or have anomalies in their definition and application. Identification of these problems has motivated us to redefine these quantitative metrics and propose a quantitative approach to determine the overall quality of a conceptual model. This paper is mainly focused on two quality attributes namely completeness and correctness, and on the metrics defined for these attributes. The study is further extended to define a quality framework which helps in computing the overall quality of a conceptual model. The results of this research provide the basis to compare two alternate models of a given problem domain and decide which model is of better quality.