Automated Method Comparison: Merging Ethical, Social, and Environmental Accounting Models Using NLP and Model-Driven Techniques
Summary
In response to the increasing prominence of ethical, social, and environmental (ESE) concerns, organizations are redefining success by prioritizing the well-being of people and the planet. The integration of ethical, social, and environmental accounting (ESEA) is gaining importance, with various methods being employed. However, despite the growing numerous of ESEA methods available, there is a significant overlap in the indicators they require allowing many organizations end up utilizing multiple ESEA methods and having to input the same data into disconnected tools. This paper introduces an innovative automated method comparison algorithm designed to identify relationships, calculate similarities, and merge ESEA methods. The primary objective is to integrate this automated method comparison into the openESEA framework, which was created to streamline ESEA methods but currently lacks an automated solution for assessing similarities and merging them. Utilizing the Design Cycle framework, which encompasses problem investigation, treatment design, and treatment validation phases, the research addresses seven sub-research questions. A comprehensive multivocal literature review explores existing approaches for assessing similarities and differences between ESEA methods, incorporating investigations into model management and merging strategies. The treatment design phase involves crafting requirements through user stories, expanding the domain-specific language (DSL), and developing an automated method comparison algorithm using natural language processing techniques, similarity measures, and large language models. To validate the design, a gold standard experiment and algorithm assessment verification were conducted. The findings reveal variability in precision values, emphasizing the need for a nuanced exploration of precision-recall trade-offs. Despite challenges, the automated approach offers practical benefits by saving time and reducing redundancy. This research establishes a foundation for innovation in matching and merging ESEA methods, offering broader applicability across domains and providing valuable insights for both academia and practitioners.