Search for a command to run...
The article is an extended version of the poster published in ISD 2024 conference: S. J. Niepostyn and W. B. Daszczuk, “An Objectified Entropy-Based Software Metric for Assessing the Maturity of Use Case Diagrams,” in 32nd International Conference on Information Systems Development, ISD2024, Gdańsk, Poland, 26–28 Aug 2024, 2024, pp. 1–5. https://doi.org/10.62036/ISD.2024.14 The main extensions are: — The method of determining weights in relation to the Abran method has been described in detail: – We discuss the difference between our method and the Abran method. – We describe the details of the previous entropy-based metrics and their importance. – We discuss the differences between the old metric FBS16 and present FBS24 (FBS16 was used solely to determine the completeness and consistency of software architecture), while FBS24 to measure information content of UML diagrams. – We describe the weaknesses of FBS16 compared to FBS24. — We detail Step 1: constructing artificial diagram series and presenting the calculation results. — We detail Step 2: description of individual diagrams and presenting their calculation results. — We discuss mature and immature diagrams using examples of specific UC diagram types (business, system, internal, implementation). — We discuss determining individual UCD diagram types (business, system, internal, implementation, actor) based on the FBS24 measure. — We add a discussion of observing FBS24 results on industrial diagrams. — We enhance the conclusions. — We expand the literature review with significant articles. In previous work, we proposed a method for evaluating UML diagrams and entire project layers (business, system, and implementation) in the Enchanted Consistent Model-Driven Architecture (e-CMDA) using the FBS (Functionality, Behavior, and Structure) metric based on normalized entropy. This metric quantifies the information content in a given dimension, with a maximum value of 1 indicating no content. The result supports decisions about diagram consistency or completeness. In this article, we apply the FBS metric to assess the functional information content of use case diagrams. The metric is computed by summing the counts of occurrences of element kinds multiplied by assigned weights, followed by calculating the normalized entropy of the resulting values. To enhance discrimination between diagrams, we propose selecting weights that produce the fastest decline in the metric relative to the number of elements. We address anomalies such as false lack of information or non-monotonicity by generating artificial diagram series with increasing complexity, testing various weight configurations, and selecting those that minimize distortion while maximizing differentiation. These configurations are then validated with real-world diagrams. Finally, we suggest threshold values to distinguish mature diagrams from immature ones, while allowing users to define their own. Additional applications, such as diagram classification, are also discussed.