Development of a Novel Integrated Ontology-Based ESG Assessment Tool with AI Assistance for SMEs

Abstract

<p>Existing environmental, social, and governance (ESG) standards and guidelines are difficult to assess and compare the ESG performance without a standardized and integrated framework. <a>A comprehensive assessment tool is required to integrate these standards and guidelines into a consistent list for assessment; this is especially true for SMEs. </a>The problem arises as to how these ESG standards and guidelines can be connected and interrelated to each other. This paper therefore aims to review the existing ESG standards and guidelines, and to propose an integrated ontology-based ESG assessment tool. The research method involves empirical review, content analysis and pilot testing. The proposed instrument includes three modules of the UN Sustainable Development Group (UNSDG) model, an UNSDG maturity model, and an ESG ontology A questionnaire and a rule-based AI recommendation assistant are developed. The instrument’s validity and reliability testing are done with five companies (30 samples). The results indicate that the content validity has values of 0.909 for S-CVR, 0.964 for S-CVI/UA, and 0.982 for S-CVI/Ave. The construct validity has a Pearson’s R of 0.7737, convergent validity of 0.983, divergent validity of −0.1, and construct validity of 0.883. The face validity has a Pearson’s R of 0.7147 at a confidence level of 95%. The internal reliability has an alpha value of 0.902. Pre-post reliability testing of the instrument shows no significant difference between the two periods (r=0.965, p<0.05, N=30), indicating that the instrument is well designed. The future assessment of the instrument will be extended to 30 company cases and 77 industries.</p>