A Topic Modeling Assessment of Emerging Research Trends in the Environmental Science and Engineering Discipline

Advancements in environmental science and engineering (ESE) research is needed towards ensuring an environmentally conscious society. Understanding fundamental research developments in the ESE discipline, can help stimulate improved collaboration and communication of nascent environmental problems. Hence, in this thesis, the author applies topic modeling analysis on 3072 abstracts collected from academic journals that publish subject matter related to ESE research from 2005-2019. Accordingly, the author applies a latent Dirichlet allocation (LDA) model on abstract metadata to infer 20 trending topics. Namely, topics on environmental impact assessments, waste management, and lead pollution. Moreover, whilst quantifying the trends at the regional level, it has been observed that countries display clearly distinguishable patterns. Thus, suggesting that ESE research communities from different countries tend to specialize in various sub-fields. Environmental scientists, environmental engineers, and journal editors (among other interested parties) will benefit from the results in terms of identifying promising topics for research collaborations.