Scholarlink

<p>Artificial intelligence (AI) has proliferated across all sectors of society: In economic and business domains, it can be used to forecast price changes, assess risk, and optimize business practices, among much more. Despite boosting performance, the use of AI tools in high-stakes contexts has clarified the need for increased model transparency. In the academic community, this need has greatly motivated research which interrogates and develops the interpretability and explainability of AI systems. This paper surveys the literature to provide a comprehensive overview of explainable and interpretable AI applied in economic and business domains. Our analysis establishes the most widely used models and techniques, the common challenges addressed, as well as the most receptive outlets for publication. A key component of our analysis is the identification of 15 main research macro-areas. Our survey provides practitioners with a detailed overview of the active research areas and, within them, research trends, illuminating avenues for future contributions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A bibliometric review of interpretability and explainability methods for artificial intelligence in economics and business

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月17日

查看全文
  • Amelia Kunze,
  • Davide La Torre,
  • Matteo Rocca

摘要

Artificial intelligence (AI) has proliferated across all sectors of society: In economic and business domains, it can be used to forecast price changes, assess risk, and optimize business practices, among much more. Despite boosting performance, the use of AI tools in high-stakes contexts has clarified the need for increased model transparency. In the academic community, this need has greatly motivated research which interrogates and develops the interpretability and explainability of AI systems. This paper surveys the literature to provide a comprehensive overview of explainable and interpretable AI applied in economic and business domains. Our analysis establishes the most widely used models and techniques, the common challenges addressed, as well as the most receptive outlets for publication. A key component of our analysis is the identification of 15 main research macro-areas. Our survey provides practitioners with a detailed overview of the active research areas and, within them, research trends, illuminating avenues for future contributions.