[论文解读] A Systematic Review of Green AI

Roberto Verdecchia, June Sallou+1 more|arXiv (Cornell University)|Jan 26, 2023

Green IT and Sustainability被引用 35

一句话总结

本论文对绿色人工智能进行系统性文献综述，分析了98个原始研究，以绘制该领域的趋势、方法和成熟度。

ABSTRACT

With the ever-growing adoption of AI-based systems, the carbon footprint of AI is no longer negligible. AI researchers and practitioners are therefore urged to hold themselves accountable for the carbon emissions of the AI models they design and use. This led in recent years to the appearance of researches tackling AI environmental sustainability, a field referred to as Green AI. Despite the rapid growth of interest in the topic, a comprehensive overview of Green AI research is to date still missing. To address this gap, in this paper, we present a systematic review of the Green AI literature. From the analysis of 98 primary studies, different patterns emerge. The topic experienced a considerable growth from 2020 onward. Most studies consider monitoring AI model footprint, tuning hyperparameters to improve model sustainability, or benchmarking models. A mix of position papers, observational studies, and solution papers are present. Most papers focus on the training phase, are algorithm-agnostic or study neural networks, and use image data. Laboratory experiments are the most common research strategy. Reported Green AI energy savings go up to 115%, with savings over 50% being rather common. Industrial parties are involved in Green AI studies, albeit most target academic readers. Green AI tool provisioning is scarce. As a conclusion, the Green AI research field results to have reached a considerable level of maturity. Therefore, from this review emerges that the time is suitable to adopt other Green AI research strategies, and port the numerous promising academic results to industrial practice.

研究动机与目标

• 理解绿色人工智能研究的特征与成熟度。
• 识别绿色人工智能中研究的主要主题、产物和阶段。
• 评估绿色人工智能研究中的方法学、行业参与和工具提供情况。

提出的方法

• 使用面向标题查询在 Google Scholar、Scopus 和 Web of Science 上进行自动化检索。
• 对预定义的纳入/排除标准进行人工应用以筛选原始研究（I1-I4；E1-E6）。
• 双向滚雪球法（向后和向前）以达到理论饱和。
• 两阶段数据提取，构建结构化框架（定义、研究类型、主题、领域、数据类型、产物、阶段、策略、数据集大小、能耗节省、行业参与、目标读者、工具可用性）。
• 使用常数/开放编码的数据整合以统一提取的概念。

实验结果

研究问题

• RQ1 绿色人工智能前沿研究的特征是什么？（总体格局与定义。）
• RQ2 绿色人工智能文献中的出版趋势、主题、领域和产物有哪些特征？
• RQ3 绿色人工智能研究中的行业参与程度和实际工具提供情况如何？
• RQ4 在绿色人工智能中通常研究的数据类型、算法和数据集规模是什么？

主要发现

• 回顾涵盖98项原始绿色人工智能研究，领域自2020年以来呈现显著增长（自2020年起发表的论文占比为76%）。
• 能耗节省最高达到115%，50%及以上的节省相对常见。
• 大多数研究聚焦在训练阶段和神经网络上，图像数据是最常使用的数据类型。
• 行业参与存在但受限（约23%），绿色人工智能的工具提供也较少。
• 监控、超参数调优、模型基准测试和部署是主导主题，而数据为中心和排放相关主题代表性不足。
• 该领域处于成熟阶段，适合将学术成果转化为工业实践。