Using fuzzy logic and neural networks to classify socially responsible organisations

Academics and practitioners have not yet developed an adequate method to evaluate the social performance of organisations that includes a robust and comprehensive approach of sustainability and uses the most relevant data sources. However, sustainability rating agencies are evaluating the social performance of organisations according to their own methodologies, which are not always clearly explained to stakeholders; and the evaluations they provide are being used as a reference in markets. This study contributes to research on the evaluation of social performance in organisations, by means of an innovative methodology that combines the use of neural networks and fuzzy logic for the development of expert systems suitable for classifying organisations according to their performance on Corporate Social Responsibility. The methodology has been validated in a simplified scenario and results indicate that it is suitable for replicating the classifications provided by sustainability rating agencies.