Debatt ● Jerry Chun-Wei Lin
Unjustified Criticism in Khrono
My case will be handled by the National Committee for the Investigation of Research Misconduct. But since the criticism has been made public, I will comment on some of the errors that characterize the presentations in Khrono.
Publisert
Sist oppdatert
Denne teksten er et debattinnlegg. Innholdet i teksten uttrykker forfatterens egen mening.
Khrono presented on 12 and 16 September 2025 two cases related to Western Norway University of Applied Sciences' (HVL) decision regarding alleged scientific misconduct attributed to Jerry Chun-Wei Lin (JCWL). We (JCWL and his representative) provided detailed responses to several questions from Khrono before the first case was published. We expected those responses to be included in the case, but many of them were ignored. Consequently, important context concerning the discrepancies between claims from HVL and JCWL was omitted, leaving readers without crucial information.
Initially, it should be noted that a high number of published articles is not evidence in itself. The number of published articles is high by Norwegian standards, but is more common for researchers in a number of other countries.
The specific case will be handled by the National Committee for the Investigation of Research Misconduct (Investigation Committee)/ Nasjonalt utvalg for gransking av uredelighet i forskning (Granskingsutvalget). But since the criticism has been made public, we will comment on some of the errors that characterize the presentations in Khrono.
Background: HVL's decision relies heavily on an external report that contains inaccuracies, all of which we addressed in our replies, which seem to have been dismissed by HVL when they discussed the case: «The extent and nature of the errors identified by the committee cannot be rectified by sending additional data or providing further explanations of methodology. Therefore, it is not necessary for the case's processing to conduct a detailed assessment of the supplementary material.» This dismissal raises procedural fairness concerns regarding how these claims were analyzed.
Categorization of Publications: There is a distinction between JCWL's core research publications (including his submissions and those of his PhD candidates) and those resulting from his external collaboration efforts, which often stem from master's projects conducted at the labs of collaborators. His core research papers have been unjustly criticized for their quality and alleged nonsensical text. However, the examples cited as nonsensical indicate that these criticisms are based on a lack of understanding of domain-specific terminology, as seen below. The high number of coauthors arises from master’s projects requiring publication in certain jurisdictions. Criticisms of use cases like «apply model X to application Y» (often the case for a master’s project, can be of interest for practitioners) overlook the complexity involved in such adaptations. Assertions lacking improvements over time are unfounded, given the distinct groups of authors for the papers examined. It's important to note that the papers in question were published before the advent of tools like ChatGPT, which could have aided in refining language and content.
Domain-Specific Knowledge Requirement: Quality evaluations must be undertaken by reviewers with domain-specific expertise. Misinterpretations of domain-specific terminology as nonsense indicate inadequate understanding, as demonstrated by the HVL librarian's critique during a Khrono interview. The term «prevention of cybersecurity/cybersecurity prevention» was incorrectly dismissed, despite being a well-defined phrase in the field that covers strategies to prevent fraud, attacks, and risks—not cybersecurity itself. Such misconceptions, which can unjustly impact perceptions of a paper's quality, could have been avoided with a basic literature search through google scholar, which would reveal that it is a commonly used concept in the field, as evidenced by courses offered by reputable institutions like San Diego State University. Understanding domain-specific language is crucial, and even experts must carefully investigate terminology outside their subfields to ensure accurate evaluations.
Assessment of Level of Criticisms: In the external report, many of the papers received brief comments that we consider rather mild. One example of nonsensical text was the use of «single value decomposition» instead of the correct «singular value decomposition». This is a common mistake across scholarly articles and does not warrant severe judgment on the overall quality of the paper. Code Sharing Critique: The report questioned the validity of some papers due to the absence of shared code, implying potential fabrication. It is crucial to note that the journals did not require code sharing. While code and patents were eventually provided to HVL, this evidence was dismissed because it was presented after the critique.
Peer Review Integrity Speculations: The report's speculation regarding corrupted peer review processes is unsubstantiated. We intend to provide documentation during our appeal to Granskingsutvalget, illustrating that thorough standard peer reviews by domain-specific experts were conducted.
Summary of 2021 Meeting Discrepancies and Case Handling Issues: Significant discrepancies exist between JCWL's and HVL's accounts of a 2021 meeting. JCWL recalls attending a meeting without directives to reduce publications. HVL claims otherwise. Evidence from 2021 meeting minutes that was received in 2025 supports JCWL's recollection, noting only guidelines but no directive on reducing publication numbers. Documentation was not shared post-meeting, failing procedural standards. Similarly, case handling in 2023 exhibited procedural issues, with HVL not sharing meeting minutes or faculty reports with involved parties, affecting their ability to address inaccuracies. Regulatory violations are evident in HVL's documentation handling, exemplifying systemic procedural issues that merit further examination.
Conclusion: The criticisms leveled are largely based on misunderstandings of academic terms, domain-specific language, and editorial choices. Code and patents were provided post-publication to counter allegations of fabrication. The peer review process was unjustly criticized. We advocate for acknowledgment of our corrections and a thorough investigation into HVL's decision-making process and case handling practices. This situation underscores broader procedural challenges faced by HVL that require redress.
