An Ontology for Software Patterns: Application to Blockchain-Based Software Development.

Ensuring the quality of software design is usually a difficult task. In the blockchain field, the design of an application is particularly important as flaws can lead to critical vulnerabilities and cost overheads. To assist practitioners in this task, software patterns can be used (solutions to repeatable problems in a given context). Some blockchain patterns exist but they are scattered, and described in many different notations and templates. As a result, practitioners can be lost in the selection of adequate blockchain-based patterns. This paper fills the gap by proposing a blockchain-based software pattern ontology. The ontology is composed of two distinct subspaces: first, a set of classes and individuals related to blockchain-based patterns, based on a previous Systematic Literature Review (SLR). It notably reuses a taxonomy of blockchain design patterns, that helps to classify these patterns in relevant categories. Along that, another subspace has been created to further organize the knowledge related to software patterns and allow inferences. A tool is proposed along with the ontology to assist practitioners in finding blockchain-based patterns that fit their needs. An evaluation is performed to assess the usability and the relevancy of the ontology.