Academic Fraud Detection Report: Optimization of Sorting Operations at Express Delivery Distribution Centers (DOI: 10.3969/j.issn.1007-5429.2021.03.001)

Academic fraud report · Geng Detector

Summary

Verdict: Questionable. This report examines a 2021 paper by Zhang Yang, Ma Ruyi, Liu Congyu, Li Cuiduo, and Jiang Zehui, published in Industrial Engineering and Management (Vol. 26, No. 3), which uses FlexSim simulation to optimize sorting operations at express delivery distribution centers. Key concerns: (1) Table 6 reports suspiciously narrow efficiency ranges of 2.34% and 2.03% across six operators under the proposed dynamic combination algorithm, with average efficiencies near 94.85% and 94.62%—statistically improbable given the exponential arrival distribution and highly skewed destination proportions stated in the model. (2) Table 3 claims randomly generated destination proportions yet lists suspiciously tidy two-digit values summing exactly to 1.00, inconsistent with typical pseudorandom output. (3) Severe encoding corruption (UTF-8 decoded as GBK/GB2312) throughout the source document impedes full verification. Image forensics could not be performed due to lack of decodable figures. Confidence in the statistical anomaly findings is moderate; image-based fraud could not be ruled in or out. Final judgment requires raw FlexSim model files and run logs.

Verdict

Questionable (🟡). The paper shows statistical anomalies in simulation outputs that warrant verification, but cannot be definitively classified as fraud without the original FlexSim model files, run logs, and full image access.

Key findings

Suspiciously uniform simulation outputs: Table 6 reports operator efficiency ranges of only 2.34% and 2.03% under the proposed dynamic combination algorithm, with average efficiencies of 94.85% and 94.62%. Given the model uses an exponential arrival distribution and highly skewed destination shares, such uniformity is statistically improbable.
Implausible "random" inputs: Table 3 lists ten destination proportions as 0.19, 0.15, 0.14, 0.12, 0.12, 0.10, 0.08, 0.06, 0.03, 0.01, summing exactly to 1.00 with two-decimal precision—atypical for true pseudorandom output.
Severe text encoding corruption: The source document contains widespread mojibake (e.g., µÚ, ¾í, ÆÚ) consistent with UTF-8 misinterpreted as GBK/GB2312, hampering full-text verification.
Methodological packaging concerns: The proposed algorithms resemble classical bin-packing / line-balancing heuristics (greedy pairing of high- and low-volume destinations), raising novelty questions but not constituting fraud.
Image forensics not performed: No usable figure data could be extracted from the corrupted source.

Evidence highlights

DOI: 10.3969/j.issn.1007-5429.2021.03.001
Journal: *Industrial Engineering and Management* (工业工程与管理), Vol. 26, No. 3, 2021
Table 6 efficiency ranges: 2.34%, 2.03%
Table 6 average efficiencies: 94.85%, 94.62%
Table 3 destination proportions: 0.19, 0.15, 0.14, 0.12, 0.12, 0.10, 0.08, 0.06, 0.03, 0.01 (sum = 1.00)
Stated arrival process: Exponential distribution

Notes

Authors should be asked to provide the original FlexSim model files and simulation run logs to verify whether results were cherry-picked or parameters manually tuned.
The encoding corruption is a serious technical concern regardless of authorship intent; the journal's digital publishing pipeline may warrant audit.
All numeric values above are preserved exactly as reported in the source detection report.
No image manipulation conclusions are drawn due to insufficient figure data.