Generative AI in Soft Computing: Methodological Implications and a Literature Survey on MCDM and Type-2 Neutrosophic Applications

Abstract

This literature survey presents a systematic review of existing research on weighting and ranking MCDM methods and their application in Generative AI. This review highlights a significant gap: few GenAI studies use MCDM in a type 2 neutrosophic environment to rank BC-integrated LLMs and GenAI chatbots, despite substantial theoretical development. Two innovative applied frameworks are being introduced to close this gap: T2NN-RANCOM-MARCOS for the assessment of GenAI chatbots and T2NN-CRITIC-MAIRCA for the ranking of blockchain based on LLMs. Additionally, applying the comparative and sensitivity analyses on two methodologies, which should show how integrating Type-2 neutrosophic number can produce more stable, uncertainty-aware rankings. Developing capabilities that (i) allow multiple subject-matter experts to provide judgments linguistically via Type-2 Neutrosophic Numbers (T2NNs), thereby eliminating ranking inconsistencies caused by extreme expert data values; and (ii) calculate experts' weights in the decision-making process by proposing a novel weighting approach that assesses an expert's skill level through linguistic T2NN information. Providing brief explanations of the fundamental techniques included Type-2 Neutrosophic Number (T2NN), CRITIC, MAIRCA, RANCOM, and MARCOS. Future research in decision-making for intelligent systems under ambiguity is anticipated to be guided by the survey.