On the use of the analytic hierarchy process in the evaluation of domain-specific modeling languages for multi-agent systems


Asici T. Z. , Tezel B. T. , KARDAŞ G.

JOURNAL OF COMPUTER LANGUAGES, vol.62, 2021 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 62
  • Publication Date: 2021
  • Doi Number: 10.1016/j.cola.2020.101020
  • Title of Journal : JOURNAL OF COMPUTER LANGUAGES

Abstract

Software agents and Multi-agent Systems (MAS) composed by these agents are used in the development of the complex intelligent systems. In order to facilitate MAS software development, various domain-specific modeling languages (DSMLs) exist. Unfortunately, the usability evaluation of these languages are mostly not considered or only a few assessments which cover one single MAS DSML are made. A comparative evaluation, which is missing in the existing studies, may help agent software developers to choose the MAS DSML which fits well into the system development requirements. Hence, in this paper, we introduce a comparative MAS DSML evaluation methodology based on the Analytical Hierarchy Process (AHP). A categorized set of MAS DSML criteria which can be used for the multi-criteria decision making is defined. These criteria can be prioritized by the developers according to their modeling language expectations and the application of the methodology allows the evaluation of DSML alternatives based on this prioritization. As the result of the automatic calculation of the importance distributions, the most appropriate DSML is determined. With the voluntarily participation of a group of agent software developers, the proposed methodology was applied for the comparative evaluation of four well-known MAS DSMLs. The conducted evaluation showed that the agent developers prioritized appropriateness, completeness and shortening the development time as the most significant criteria for the MAS DSML assessment while the attractiveness of the notations had a minimum effect on preferring a language. Favorite DSML for each comparison category and criteria was determined within this evaluation.