Scholarlink

<p>This paper aims to explore trends in the application of big data and Machine Learning (ML) in Water Resources Management (WRM) by categorizing research studies into distinct scientific subfields. A comprehensive analysis was performed on articles published between 2018 and 2024. Leveraging a dataset of 6,430 collected papers, 173 articles were evaluated using bibliometric techniques to track the development of academic interest and recognize pivotal studies. Our suggested unsupervised classification model established categories and organized relevant articles according to their specific scientific focus, using keywords extracted from titles, abstracts, and author-defined keywords, with stop-words excluded. The model achieved a validation accuracy of 90.16% through the Multinomial Naïve Bayesian (MNB), 86.54% Random Forest (RF) and 84.61% Support Vector Machine (SVM) approaches. The analysis revealed 10 distinct research topics, emphasizing WRM and innovative city applications as leading categories. This study contributes to the literature by introducing a methodology for analyzing existing research, uncovering emerging scientific areas within WRM, and highlighting potential avenues for future investigation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Systematic review and topic classification of soft computing and machine learning in water resources management

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月25日

查看全文
  • Maria Drogkoula,
  • Nicholas Samaras,
  • Omiros Iatrellis,
  • Eftihia Nathanail,
  • Konstantinos Kokkinos

摘要

This paper aims to explore trends in the application of big data and Machine Learning (ML) in Water Resources Management (WRM) by categorizing research studies into distinct scientific subfields. A comprehensive analysis was performed on articles published between 2018 and 2024. Leveraging a dataset of 6,430 collected papers, 173 articles were evaluated using bibliometric techniques to track the development of academic interest and recognize pivotal studies. Our suggested unsupervised classification model established categories and organized relevant articles according to their specific scientific focus, using keywords extracted from titles, abstracts, and author-defined keywords, with stop-words excluded. The model achieved a validation accuracy of 90.16% through the Multinomial Naïve Bayesian (MNB), 86.54% Random Forest (RF) and 84.61% Support Vector Machine (SVM) approaches. The analysis revealed 10 distinct research topics, emphasizing WRM and innovative city applications as leading categories. This study contributes to the literature by introducing a methodology for analyzing existing research, uncovering emerging scientific areas within WRM, and highlighting potential avenues for future investigation.