Integrity review of 'Attribute-based Privacy-Preserving Encryption Scheme with Inner Product Predicate' (DOI: 10.11999/JEIT221050)

Academic fraud report · Geng Detector

Summary

This report evaluates the theoretical cryptography paper by Zhang Zhiqiang, Zhu Youwen, Wang Jian, and Zhang Yushu, published in Journal of Electronics & Information Technology (2023), proposing an attribute-based encryption scheme with inner-product predicates. The reviewer concludes the paper is likely clean (verdict: cleared), based on textual and logical analysis only. Key checks: theoretical complexity claims in Table 3 were verified by arithmetic — (12n+3)E − (4n+2)E = 8n+1E versus Wang Yue et al. and (7n+4)E − (4n+2)E = 3n+2E versus Zhang et al., both matching the paper's stated savings exactly. The six algorithms defined in Section 3.3 align one-to-one with the construction steps in Section 4, with no internal n-value contradictions. The submission timeline (received 2022-08-10, revised 2022-11-01, online 2022-11-05) is temporally consistent. Limitations: only text was provided, so pixel-level image forensics on Figures 1–4 (system model and performance plots) were not performed; duplication or smoothing artifacts in the line charts cannot be ruled out. Confidence is moderate-to-high for textual/mathematical integrity, low for visual integrity.

Verdict

Cleared based on available textual and mathematical evidence. No substantive integrity issues detected. A residual caveat applies to image-based checks, which were not possible from the supplied text-only input.

Key findings

Internal mathematical self-consistency verified. Two explicit complexity-reduction claims in Table 3 were checked by direct subtraction of the reported exponentiation counts and both match exactly.
Algorithmic structure is consistent. The six algorithms defined in Section 3.3 map one-to-one with the six construction steps in Section 4; no parameter drift in n.
Citation set is plausible. Cited works (e.g., Katz et al. [7], Lai et al. [12]) are standard references in predicate/attribute-based encryption and align with the described security arguments.
Timeline is internally coherent. Received 2022-08-10, revised 2022-11-01, online 2022-11-05; no anachronistic references.
Image integrity unverifiable from input. Figures 1–4 (system model and four performance line charts) could not be subjected to pixel-level duplication or PS-trace analysis because only extracted text was provided.

Evidence highlights

Claim vs. computation, Table 3, encryption cost:
Paper states savings of 8n+1 exponentiations vs. Wang Yue et al. [10].
Reported values: Wang [10] = (12n+3)E; proposed PPES = (4n+2)E.
Difference: (12n+3) − (4n+2) = 8n+1 ✅ matches.
Claim vs. computation, Table 3, encryption cost:
Paper states savings of 3n+2 exponentiations vs. Zhang et al. [11].
Reported values: Zhang [11] = (7n+4)E; proposed PPES = (4n+2)E.
Difference: (7n+4) − (4n+2) = 3n+2 ✅ matches.
Section 3.3 algorithm count (6) equals construction-step count in Section 4 (6); no contradictory n across sections.
DOI 10.11999/JEIT221050 preserved as supplied.

Notes

Scope of analysis is limited to text extracted from the PDF; original high-resolution figures were not available, so visual forensic checks (image reuse, splicing, smoothing artifacts in Figures 2–4) remain an open gap.
Because the experimental "data" in this domain are theoretical complexity counts rather than measured quantities, traditional statistical-fraud indicators (e.g., fabricated instrument readings) are not applicable.
The verdict of "cleared" reflects the text-based checks performed; a definitive determination would require access to source figures and, ideally, reviewer reports from the journal.