In a wide range of areas in information technology (IT), the reusability of previously created artifacts is an essential and critical domain of problems. This paper focuses on web engineering as a specific branch of software engineering and systems modeling. Web engineering has similarities and dissimilarities to other engineering disciplines within IT. The model-driven approach can be considered as a common development methodological approach with other engineering disciplines; however, web engineering has its own specialties. As the paper proposes, well-known models based on the unified modeling language (UML) could be used for the reusability of models, namely the use case diagram and activity diagram. In a web engineering environment, it is self-evident to exploit the opportunities offered by the semantic web.

The basic idea of the proposal is that the results/models of previous projects should be stored in an appropriate semantic representation. When starting new projects, the functional requirements conceptualize in the form of use case diagrams, and then an algorithm semi-automatically generates activity diagrams with annotation. To create the annotations for activity diagrams, the algorithm needs a repository that contains the semantic rich description of earlier models.

A novel algorithm has been worked out for measuring similarity between use case diagrams. The algorithm makes use of WordNet, and a part-of-speech (POS) tagger. Based on the similarities between use case diagrams of functional requirements and previous models, a proposal for reusing earlier models is generated. As a proof of concept, an empirical study and statistical analysis were carried out. The paper’s contributions are: a similarity metric and algorithm for use case diagrams, an algorithm for annotations of activity diagrams, and the customization of activity diagrams.

The paper is interesting for researchers working in various information systems modeling and software engineering domains who would like to exploit the potential of semantic technologies in those fields.