Abstract:
Even though author position is taken into account by some scores, they ignore
actual contributions, particularly for middle authors or for alphabetical ordering.
Furthermore, position-based scores frequently contain over-rewards and redundancies,
like the single-author factor being repeated across sub-scores. Citations
differ by domain and year of publication as well, but the existing scores do not
account for multi-domain involvement or outlier effects, such as a few papers with
a lot of citations or a few with very few. Likewise, h, g, Hf, and Hm indices vary
according to article domains and career stage, which disadvantages early-career
academics in sectors with few citations. Lastly, while Hm and Hf assume equal
author contributions, the m-index penalizes seniors. To solve these inadequacies,
this work introduces the Nilmantha (Nm) index, a novel percentile-based metric
designed to address these limitations through comprehensive normalization and
weighting strategies that produce a percentile score showing academic significance
directly comparable internationally. Alphabetical/Random (A/R), Relative Contribution
(RC) with multiple/single First Authors (FAs) or Last Authors (LAs),
and Explicit Contribution (EC) with multiple/single FAs/LAs are considered. In
particular, it employs a harmonic model for RC, uniform weights for A/R, proportionate
uniform weights for multiple FAs/LAs per RC/EC, and AHP-driven
CRediT or ACI. First, self-influence and overlapping core-author influence are
taken into account when adjusting citations for influential self-citations. Second,
we calculate the weighted average hf (productive cited output with fractional
credit), hm (productive cited output for multi-authorship), and g (highly cited
works) indices, all of which are adjusted for career factor (preserving senior academics’
collective achievements) and multiple fields. We also calculate the outlieruncontrolled
weighted field- and year-normalized (WFYN) total cites (overall impact),
outlier-controlled WFYN citations per paper (average paper impact), and
outlier-controlled WFYN citation rate (impact rate). Fictitious author profile
use cases are used to validate the model, and the numerical results show that
the model captures 36% more impact than traditional C-scores in multi-field contexts,
provides 118% more accurate per-paper assessments through contribution
weighting, and avoids the over-penalization of senior researchers while appropriately
boosting early-career scholars.
