A Roadmap of Emerging Trends Discovery in Hydrology: A Topic Modeling Approach

TL;DR

This study uses topic modeling techniques to identify and analyze emerging trends in hydrology research, highlighting key topics like climate change and water management that persist over recent years.

Contribution

It compares multiple topic models to effectively discover and validate trending research topics in hydrology using coherence scores.

Findings

Identified key trending topics in 2022 and 2023

Compared effectiveness of LDA, NMF, and LSA models

Confirmed stability of main topics over two years

Abstract

In the new global era, determining trends can play an important role in guiding researchers, scientists, and agencies. The main faced challenge is to track the emerging topics among the stacked publications. Therefore, any study done to propose the trend topics in a field to foresee upcoming subjects is crucial. In the current study, the trend topics in the field of "Hydrology" have been attempted to evaluate. To do so, the model is composed of three key components: a gathering of data, preprocessing of the article's significant features, and determining trend topics. Various topic models including Latent Dirichlet Allocation (LDA), Non-negative Matrix Factorization (NMF), and Latent Semantic Analysis (LSA) have been implemented. Comparing the obtained results with respect to the $C_V$ coherence score, in 2022, the topics of "Climate change", "River basin", "Water management", "Natural hazards/erosion", and "Hydrologic cycle" have been obtained. According to a further analysis, it is shown that these topics keep their impact on the field in 2023, as well.