文献类型：期刊文献

英文题名：Multi-rung intuitionistic fuzzy group decision-making with golden-mode data center for AI-based software quality evaluation

作者：Liu, Taoli[1];Yue, Chuan[1]

机构：[1]Guangdong Ocean Univ, Coll Math & Comp Sci, Haida Rd 1st, Zhanjiang 524088, Guangdong, Peoples R China

年份：2026

卷号：17

期号：10

外文期刊名：AIN SHAMS ENGINEERING JOURNAL

收录：SCI-EXPANDED(收录号：WOS:001844516800001)、、EI(收录号：20263221255722)、Scopus(收录号：2-s2.0-105046460438)、WOS

基金：The authors gratefully acknowledge the financial support pro vided by the National Key R&D Program of China (Grant No. 2024YFC3308004) .

语种：英文

外文关键词：Artificial intelligence; Software quality evaluation; Group decision-making; Intuitionistic fuzzy number; Linguistic variable; Entropy measure

外文摘要：Evaluating AI-based software quality requires robust methods to synthesize uncertain, multi-expert judgments. This study develops a novel group decision-making framework within a multi-rung intuitionistic fuzzy envi ronment. Unlike existing methods that rely on mean-based consensus anchoring and ad hoc expert weight assignments, the proposed framework introduces three interconnected innovations. First, a golden-mode data center is proposed for robust consensus anchoring, which reduces sensitivity to outlier judgments compared to conventional mean-based centers. Second, an entropy-based measure is adapted to quantify expert data quality objectively, replacing subjective weight assignments. Third, a systematic aggregation process integrates expert weights, attribute weights, and group consensus into a unified decision pipeline. Experimental validation using real-world AI-based software quality data demonstrates the framework's superiority. The golden-mode data cen ter achieves a 10% improvement in coefficient of variation and a 12.5% improvement in weight dispersion for expert weights, along with a 60% improvement in minimum separation for alternative rankings compared to the mean-based center. Comprehensive sensitivity analysis confirms 100% ranking stability for the best and worst alternatives under attribute weight and expert weight perturbations. The framework provides a mathematically rigorous and highly stable tool for AI-based software quality evaluation, with broader implications for service system optimization and decision science under uncertainty.