We use topic modeling to study research articles in environmental and resource economics journals in the period 2000–2019. Topic modeling based on machine learning allows us to identify and track latent topics in the literature over time and across journals, and further to study the role of different journals in different topics and the changing emphasis on topics in different journals. The most prevalent topics in environmental and resource economics research in this period are growth and sustainable development and theory and methodology. Topics on climate change and energy economics have emerged with the strongest upward trends. When we look at our results across journals, we see that journals have different topical profiles and that many topics mainly appear in one or a few selected journals. Further investigation reveal latent semantic structures across research themes that only the insider would be aware.

我们使用主题建模来研究2000-2019年间环境和资源经济学期刊上的研究文章。基于机器学习的主题建模使我们能够随着时间的推移和跨期刊识别和跟踪文献中的潜在主题，并进一步研究不同期刊在不同主题中的作用以及对不同期刊中主题不断变化的重视。在此期间，环境和资源经济学研究中最普遍的主题是增长与可持续发展以及理论和方法论。有关气候变化和能源经济学的话题以上升趋势最为明显。当我们查看各期刊的结果时，我们会发现期刊具有不同的主题配置，并且许多主题主要出现在一种或几种选定的期刊中。