Scholarlink

<p>Citizen science (CS) has emerged as a collaborative process for addressing complex scientific and societal challenges. The emergence of artificial intelligence (AI) into CS projects, has transformed data collection, analysis, and validation steps. However, significant gaps remain in understanding the methodologies, applications, and challenges of AI-CS integration. Our systematic review seeks to address the gaps by answering three questions: (1) What AI methodologies are most commonly applied in CS projects? (2) How does AI integration impact the efficiency and scalability of CS initiatives? (3) What challenges arise from AI-CS integration, and how are they mitigated? Following the PRISMA-ScR guidelines, a systematic search of Scopus, ACM Digital Library, and Web of Science identified relevant articles published between 2013 and 2024. From an initial pool of 2,470 publications, 90 were retained after filtering through the eligibility criteria. Our findings illustrate ML techniques, including deep learning, clustering algorithms, and convolutional neural networks, boost data annotation, classification, and validation in applications across various disciplines. However, challenges such as data quality variability, algorithmic opacity, and scalability constraints persist. Our conclusion identifies the multifaceted role of AI in citizen science, categorized into three primary functions: (1) assisting or replacing humans in task completion, (2) influencing human behaviour and fostering engagement, and (3) improving insights through pattern identification and decision support.</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

A systematic literature review on the role of artificial intelligence in citizen science

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月22日

查看全文
  • Germain Abdul-Rahman,
  • Andrej Zwitter,
  • Noman Haleem

摘要

Citizen science (CS) has emerged as a collaborative process for addressing complex scientific and societal challenges. The emergence of artificial intelligence (AI) into CS projects, has transformed data collection, analysis, and validation steps. However, significant gaps remain in understanding the methodologies, applications, and challenges of AI-CS integration. Our systematic review seeks to address the gaps by answering three questions: (1) What AI methodologies are most commonly applied in CS projects? (2) How does AI integration impact the efficiency and scalability of CS initiatives? (3) What challenges arise from AI-CS integration, and how are they mitigated? Following the PRISMA-ScR guidelines, a systematic search of Scopus, ACM Digital Library, and Web of Science identified relevant articles published between 2013 and 2024. From an initial pool of 2,470 publications, 90 were retained after filtering through the eligibility criteria. Our findings illustrate ML techniques, including deep learning, clustering algorithms, and convolutional neural networks, boost data annotation, classification, and validation in applications across various disciplines. However, challenges such as data quality variability, algorithmic opacity, and scalability constraints persist. Our conclusion identifies the multifaceted role of AI in citizen science, categorized into three primary functions: (1) assisting or replacing humans in task completion, (2) influencing human behaviour and fostering engagement, and (3) improving insights through pattern identification and decision support.