Extent to which sneaked references distort citation counts beyond the IJISRT dataset

Determine the extent to which sneaked references—references registered in Crossref metadata but absent from the actual reference section or full text of the corresponding PDF—distort citation counts across the scholarly literature beyond the dataset from the International Journal of Innovative Science and Research Technology (IJISRT).

Background

The paper documents 80,205 sneaked references inserted into Crossref metadata for 2,782 records associated with the International Journal of Innovative Science and Research Technology (IJISRT), all benefiting the same journal. These sneaked references inflate citation counts and were shown to propagate into scientometric platforms such as Dimensions and OpenAlex.

While the paper proposes and evaluates methods (M1 and M2) to detect sneaked references at the document level and discusses an attempt to scale detection, the authors explicitly note uncertainty about the broader prevalence and impact of sneaked references outside this specific dataset. Establishing the global extent of distortion is necessary to assess the bibliometric consequences and inform mitigation strategies.

References

Beyond this specific data set, the extent to which sneaked references are distorting citation counts is unknown.

— Detection of metadata manipulations: Finding sneaked references in the scholarly literature (2501.03771 - Besançon et al., 7 Jan 2025) in Section “Future work”, Conclusions