Trustworthy Machine Learning Evaluation Framework for Robust and Interpretable Intelligent Systems

This dataset is a curated supplementary dataset for the article “Trustworthy Machine Learning Evaluation Framework for Robust and Interpretable Intelligent Systems” published in International Transactions on Artificial Intelligence (ITALIC), Vol. 4, No. 2, May 2026, pp. 148–160. The dataset structures the main evidence, framework components, methodological elements, governance criteria, benchmark dataset metadata, SDG alignment matrix, and figure/table summaries used to support the Trustworthy Machine Learning Evaluation Framework (TMLEF). The workbook includes article metadata, research questions, key AI evaluation and governance concepts, literature synthesis, data collection methods, benchmark dataset metadata, TMLEF architecture stages, governance compliance criteria, resilience enhancement cycle phases, comparative security impact indicators, and SDG alignment mapping. It is intended to support research transparency, documentation, secondary analysis, and reproducibility of the framework design. The file does not contain individual-level or sensitive records from the Adult Income, MIMIC-III, or COMPAS datasets. Instead, it provides a structured metadata and codebook-style dataset derived from the manuscript, including sample-size references and evaluation purposes reported in the article. Users seeking raw benchmark data should consult the original dataset providers and comply with their respective access, ethics, and licensing requirements.本数据集为发表于《国际人工智能汇刊（International Transactions on Artificial Intelligence，ITALIC）》2026年5月第4卷第2期、页码148–160的论文《面向鲁棒与可解释智能系统的可信机器学习评估框架》所配套的经筛选整理数据集。本数据集结构化收录了支撑可信机器学习评估框架（Trustworthy Machine Learning Evaluation Framework，简称TMLEF）的核心佐证材料、框架组成要素、方法论单元、治理准则、基准数据集元数据、可持续发展目标（Sustainable Development Goals，简称SDG）适配矩阵以及图表摘要。 本配套数据集工作簿涵盖论文元数据、研究问题、核心AI评估与治理概念、文献综述、数据收集方法、基准数据集元数据、TMLEF架构阶段、治理合规准则、韧性增强循环阶段、对比性安全影响指标以及SDG适配映射。本数据集旨在支撑研究透明度提升、成果归档、二次分析以及框架设计的可复现性。 本文件未包含成人收入数据集（Adult Income）、医疗信息集市III（MIMIC-III）以及COMPAS数据集的个体级或敏感记录。取而代之的是，本数据集提供了源自该论文的结构化元数据与代码手册风格数据集，涵盖论文中提及的样本量参考信息与评估目标。若用户需获取原始基准数据，请联系原始数据集提供商，并遵守其各自的访问、伦理与许可要求。