The Sustainable Development Goals (SDGs) suffer from a lack of national data needed for effective monitoring and implementation. Almost half of the SDG indicators are not regularly produced, and available datasets are often out-of-date. New monitoring approaches using big data are advancing rapidly and can complement official statistics to help fill critical data gaps. However, there is poor information-sharing on the latest innovations and research collaborations across different thematic areas, and limited evaluation of strengths and weaknesses for supporting national monitoring. This paper provides a systematic review of the academic literature over the past 5 years relating to the use of big data to support monitoring of the SDGs. It reviews the state-of-the-art research using big data and advanced analytics to produce new datasets, the alignment of these datasets with the official SDG indicators, the main types and sources of big data used, and the analytical methods applied. We developed a set of evaluation criteria and applied it to highlight some of the strengths and limitations of these datasets derived from big data. We find that recent research has developed a considerable range of new datasets that could contribute to monitoring 15 goals, 51 targets, and 69 indicators. Dominant focal areas of research include land and biodiversity, health, water, cities and settlements, and poverty. Satellite and Earth Observation data were the primary sources used, most commonly applied with machine learning methods and cloud computing. However, several challenges remain, including ensuring the relevance of new datasets for monitoring SDG indicators, cost and accessibility considerations, sustainability aspects, and linking global datasets to nationally owned monitoring processes.

可持续发展目标 (SDG) 缺乏有效监测和实施所需的国家数据。几乎一半的可持续发展目标指标不是定期生成的，可用的数据集通常已经过时。使用大数据的新监测方法正在迅速发展，可以补充官方统计数据，以帮助填补关键的数据空白。然而，关于不同主题领域的最新创新和研究合作的信息共享不足，对支持国家监测的优势和劣势的评估也很有限。本文对过去 5 年中有关使用大数据支持可持续发展目标监测的学术文献进行了系统回顾。它回顾了使用大数据和高级分析来生成新数据集的最新研究，这些数据集与官方可持续发展目标指标的一致性、使用的大数据的主要类型和来源以及应用的分析方法。我们制定了一套评估标准，并应用它来突出这些源自大数据的数据集的一些优势和局限性。我们发现，最近的研究开发了相当多的新数据集，可以有助于监测 15 个目标、51 个目标和 69 个指标。主要的研究重点领域包括土地和生物多样性、健康、水、城市和住区以及贫困。卫星和地球观测数据是使用的主要来源，最常用于机器学习方法和云计算。然而，仍然存在一些挑战，包括确保新数据集与可持续发展目标指标的相关性、