Bioregulatory mechanisms in ecological art research: A multidimensional analysis based on bibliometrics and machine learning predictions

<p>With the advancement of technology, human impact on the natural world has progressively intensified, leading to an increased frequency and expanded scope of environmental issues. Ecological art integrates natural and human environments, aiming to restore damaged ecosystems through artistic expression while reflecting artists’ understanding and concerns about nature. As an emerging interdisciplinary art form, ecological art has garnered significant attention; however, systematic research in this field remains insufficient. This study explores the role of bioregulatory mechanisms in ecological art, focusing on how biological materials and biochemical processes contribute to environmental health and sustainability. combining bibliometric analysis with machine learning techniques to examine ecological art literature indexed in the WOS Core Collection between 2004 and 2024, uncovering research trends, core themes, and future directions. In the bibliometric analysis, tools such as VOSviewer, CiteSpace, Bibliometrix, Pajek, and HistCite were utilized to systematically analyze co-occurrence networks, research hotspots, and their evolutionary pathways. For machine learning analysis, comparative experiments were conducted using Random Forest, Support Vector Machines (SVM), k-Nearest Neighbors (k-NN), Extreme Gradient Boosting (XGBoost), and LightGBM models. Ultimately, Random Forest was selected for predicting citation rates and interpreting influential factors. The findings indicate that machine learning techniques quantitatively reveal the centrality of “Ecological Design” and “Sustainability” as core themes in ecological art while highlighting the emergence of new topics such as “Urban Greenspace” and “Eco-Art Education” in recent years. The publication year was identified as the most critical factor influencing citation rates, followed by the number of authors and keywords. The first author’s affiliation country also significantly impacted citation performance. This study, through the integration of bibliometric and machine learning approaches, provides a novel methodological perspective for understanding the bioregulatory potential of ecological art, predicting academic influence, and guiding future scholarly and practical applications in environmental health and sustainability.</p>