Scholarlink

<p>Evaluating the impact of researchers within a scientific community remains a complex and debated issue. Traditional metrics, such as publication and citation counts, often fail to capture the multifaceted nature of scholarly influence. While several indices, such as the h-index and its variants have been proposed, they too face limitations in reflecting qualitative and longitudinal aspects of academic contributions. This study introduces a novel researcher ranking framework that integrates four distinct parameters: Normalized Citation Score (NCS), Complete Career Contribution (CCC), Temporal Citation Input (TCI), and Collaborative Prestige (CP). These parameters are combined using two ranking mechanisms: the Statistical Ranking System (SRS), based on mathematical models, and the Comprehensive Ranking System (CRS), which employs Genetic Programming (GP) to evolve domain-specific ranking formulas. The methodology was evaluated across five datasets, Computer Science, Mathematics, Neuroscience, Civil Engineering, and a combined dataset, each balanced with equal numbers of awardees and non-awardees. In SRS, models such as the Lehmer Mean and Logarithmic Mean performed effectively in highlighting awardees. In CRS, the evolved models achieved fitness values of up to 0.96 in Computer Science and 0.88 in Mathematics, with slightly lower scores in other domains. The reduced performance on the combined dataset highlights the importance of domain-sensitive modeling. The results suggest that the proposed framework offers a flexible and comprehensive approach to researcher evaluation that can adapt to domain-specific impact patterns, providing an alternative to conventional ranking metrics.</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

Beyond publication numbers: a novel approach to academic ranking using evolutionary programming

  • 期刊论文

  • 发表日期：

    2025年5月28日

查看全文
  • Ghulam Mustafa,
  • Muhammad Tanvir Afzal,
  • Abid Rauf,
  • Muhammad Abdullah Khan

摘要

Evaluating the impact of researchers within a scientific community remains a complex and debated issue. Traditional metrics, such as publication and citation counts, often fail to capture the multifaceted nature of scholarly influence. While several indices, such as the h-index and its variants have been proposed, they too face limitations in reflecting qualitative and longitudinal aspects of academic contributions. This study introduces a novel researcher ranking framework that integrates four distinct parameters: Normalized Citation Score (NCS), Complete Career Contribution (CCC), Temporal Citation Input (TCI), and Collaborative Prestige (CP). These parameters are combined using two ranking mechanisms: the Statistical Ranking System (SRS), based on mathematical models, and the Comprehensive Ranking System (CRS), which employs Genetic Programming (GP) to evolve domain-specific ranking formulas. The methodology was evaluated across five datasets, Computer Science, Mathematics, Neuroscience, Civil Engineering, and a combined dataset, each balanced with equal numbers of awardees and non-awardees. In SRS, models such as the Lehmer Mean and Logarithmic Mean performed effectively in highlighting awardees. In CRS, the evolved models achieved fitness values of up to 0.96 in Computer Science and 0.88 in Mathematics, with slightly lower scores in other domains. The reduced performance on the combined dataset highlights the importance of domain-sensitive modeling. The results suggest that the proposed framework offers a flexible and comprehensive approach to researcher evaluation that can adapt to domain-specific impact patterns, providing an alternative to conventional ranking metrics.