Formulating rules to identify key researchers in computer science : A quantitative approach
In the vast expanse of scientific literature, determining which researchers have made the most significant con- tributions to their field is essential. However, the scientific community lacks a standardized set of criteria for identifying impactful researchers. Existing platforms, like Google Scholar, Semantic Scholar & Web of Science, provide useful metrics like total publications, citation counts, and h-index, but no universally accepted frame- work exists to comprehensively evaluate a researcher’s true impact. This study proposes a novel framework for identifying impactful researchers within the computer science domain. The framework integrates 64 dis- tinct quantitative parameters across categories such as citation-based, publication-based, author-countbased, and age-weighted metrics to evaluate researchers’ contributions. Unlike traditional methods that focus on individual indicators, our framework provides a comprehensive evaluation model that considers multiple factors simulta- neously. The parameters were selected based on their importance scores, determined by their ability to classify awardees and non-awardees. Experiments were performed on a balanced dataset comprising 600 awardees and 600 non-awardees, yielding classification accuracies ranging from 57% to 78% in recognizing influential researchers. The top-ranked features from each parameter category effectively promoted 50% to 55% of awardees into the top 100 researcher rankings. These findings offer a robust tool for academic institutions and organiza- tions seeking to identify impactful researchers. The proposed framework enables more objective evaluation and provides actionable insights for individual researchers striving for recognition and influence in their field.
Authors
- Ghulam Mustafa (ORCID: https://orcid.org/0000-0002-0354-8229)
- Muhammad Tanvir Afzal (ORCID: https://orcid.org/0000-0002-9765-8815)
Publication Details
- Journal
- Collnet Journal of Scientometrics and Information Management
- Published
- 2025-01-01
- DOI
- https://doi.org/10.47974/cjsim-2024-11007
- Citations
- 2
- Primary Topic
- Online Learning and Analytics
- Type
- article
- Field-Weighted Citation Impact
- 3.39