Differentiating Texts Written by Humans and AI: Using Spherical Fuzzy Numbers for Criterion Weighting
摘要
The rapid advancement of technology has led to numerous innovations in the field of artificial intelligence (AI). The widespread adoption of AI-based tools has brought significant transformations across various domains, particularly in written content generation. Today, AI chatbots have revolutionized written content generation by producing highly comprehensive and fluent texts that closely mimic human writing, thereby blurring the line between AI-generated and human-authored content. This study has identified a range of criteria to differentiate between AI and human writing, with determined significance levels for each criterion. Factors such as grammatical consistency, contextual coherence, originality, depth of meaning, and stylistic differences have been evaluated, and a criterion weighting process has been conducted accordingly. To ensure a robust analysis, widely used AI chatbots have been considered as expert opinion sources, providing a unique perspective to the study. To address uncertainty and expert opinion variations, the study employed Spherical Fuzzy Numbers (SFN) in the criterion weighting process for a more accurate analysis. The SF-AHP method has been applied to determine the relative importance of each criterion. This approach, based on multi-criteria decision-making, enables the defining appropriate weighting values through a systematic process. The results of the study have shown that utilizing AI chatbots as expert opinion sources in conjunction with SFN and SF-AHP methods can lead to more reliable and accurate decision-making processes. By incorporating these advanced techniques, organizations can make informed choices that consider various perspectives and uncertainties in a structured manner.