Academic Journal
View in EDS

Interpretable generative machine learning platform for chronological prediction of forensic paper aging via GC-IMS volatiles characterization in judicial authentication.

Bibliographic Details
Title: Interpretable generative machine learning platform for chronological prediction of forensic paper aging via GC-IMS volatiles characterization in judicial authentication.
Authors: Lu W; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China. Electronic address: whlu@sdupsl.edu.cn., Kong L; Institute of Eco-Environmental Forensics, School of Environmental Science and Engineering, Shandong University, Qingdao 266237, China., Zhang L; Shandong Provincial Key Laboratory of Ubiquitous Intelligent Computing, University of Jinan, Jinan 250022, China., Ma J; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China., Qi Y; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China., Lei M; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China., Li X; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China.
Source: Journal of chromatography. A [J Chromatogr A] 2026 Aug 30; Vol. 1783, pp. 467149. Date of Electronic Publication: 2026 May 29.
Publication Type: Journal Article
Language: English
Journal Info: Publisher: Elsevier Country of Publication: Netherlands NLM ID: 9318488 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1873-3778 (Electronic) Linking ISSN: 00219673 NLM ISO Abbreviation: J Chromatogr A Subsets: MEDLINE
Imprint Name(s): Original Publication: Amsterdam ; New York : Elsevier, 1993-
MeSH Terms: Volatile Organic Compounds*/analysis , Gas Chromatography-Mass Spectrometry*/methods , Forensic Sciences*/methods , Ion Mobility Spectrometry*/methods , Paper* , Machine Learning*, Boosting Machine Learning Algorithms ; Humans
Abstract: The determination of paper aging is crucial for forensic document examination, enabling the identification of temporal origin in forgery or fraud cases. Paper aging is difficult to distinguish due to the high compositional complexity, intricate temporal evolution mechanisms, and challenges in tracing diagnostic markers, limiting reliable chronological authentication. Herein, an interpretable generative end-to-end machine learning platform integrating gas chromatography-ion mobility spectrometry (GC-IMS) with machine learning was established for prediction of the temporal origin of forensic paper aging. Sixty-three differential volatiles were identified in paper materials via GC-IMS analysis. Chronological classification and regression machine learning models were constructed based the volatile fingerprints from papers aged 1 to 35 years. The CatBoost model achieved the best overall performance in both age-class identification and year prediction based on characteristic volatile profiles. SHAP (SHapley Additive exPlanations) model identified key markers as the most influential predictors, such as 3‑hydroxy-2-butanone-D, Hydroxyacetone, Cyclopentanone, and Isomenthone, enabling paper samples specific attribution of prediction outcomes. An end-to-end chronological prediction software tool was developed to automate data import, age prediction, and interpretation result generation, thereby improving its practicality and accessibility for forensic applications. Overall, this study provides a deployable analytical framework for interpretable chronological prediction of forensic paper aging and is expected to generate reliable results for judicial authentication.
(Copyright © 2026 Elsevier B.V. All rights reserved.)
Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Contributed Indexing: Keywords: End-to-end generative platform; Forensic paper aging; GC-IMS; Interpretable machine learning; Volatile organic compounds
Substance Nomenclature: 0 (Volatile Organic Compounds)
Entry Date(s): Date Created: 20260602 Date Completed: 20260629 Latest Revision: 20260629
Update Code: 20260630
DOI: 10.1016/j.chroma.2026.467149
PMID: 42229280
Database: MEDLINE
FullText Text:
  Availability: 0
CustomLinks:
  – Url: https://www.doi.org/10.1016/j.chroma.2026.467149?
    Name: ScienceDirect (all content) (s7799221)
    Category: fullText
    Text: View record from ScienceDirect
    MouseOverText: View record from ScienceDirect
  – Url: https://resolver.ebsco.com/c/fiv2js/result?sid=EBSCO:cmedm&genre=article&issn=18733778&ISBN=&volume=1783&issue=&date=20260830&spage=467149&pages=467149&title=Journal of chromatography. A&atitle=Interpretable%20generative%20machine%20learning%20platform%20for%20chronological%20prediction%20of%20forensic%20paper%20aging%20via%20GC-IMS%20volatiles%20characterization%20in%20judicial%20authentication.&aulast=Lu%20W&id=DOI:10.1016/j.chroma.2026.467149
    Name: Full Text Finder (for New FTF UI) (ns324271)
    Category: fullText
    Text: Full Text Finder
    MouseOverText: Full Text Finder
Header DbId: cmedm
DbLabel: MEDLINE
An: 42229280
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Interpretable generative machine learning platform for chronological prediction of forensic paper aging via GC-IMS volatiles characterization in judicial authentication.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AU" term="%22Lu+W%22">Lu W</searchLink>; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China. Electronic address: whlu@sdupsl.edu.cn.<br /><searchLink fieldCode="AU" term="%22Kong+L%22">Kong L</searchLink>; Institute of Eco-Environmental Forensics, School of Environmental Science and Engineering, Shandong University, Qingdao 266237, China.<br /><searchLink fieldCode="AU" term="%22Zhang+L%22">Zhang L</searchLink>; Shandong Provincial Key Laboratory of Ubiquitous Intelligent Computing, University of Jinan, Jinan 250022, China.<br /><searchLink fieldCode="AU" term="%22Ma+J%22">Ma J</searchLink>; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China.<br /><searchLink fieldCode="AU" term="%22Qi+Y%22">Qi Y</searchLink>; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China.<br /><searchLink fieldCode="AU" term="%22Lei+M%22">Lei M</searchLink>; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China.<br /><searchLink fieldCode="AU" term="%22Li+X%22">Li X</searchLink>; Characteristic Laboratory of Forensic Science in Universities of Shandong Province, Shandong University of Political Science and Law, Shandong Province, Jinan 250014, PR China.
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%229318488%22">Journal of chromatography. A</searchLink> [J Chromatogr A] 2026 Aug 30; Vol. 1783, pp. 467149. <i>Date of Electronic Publication: </i>2026 May 29.
– Name: TypePub
  Label: Publication Type
  Group: TypPub
  Data: Journal Article
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: TitleSource
  Label: Journal Info
  Group: Src
  Data: <i>Publisher: </i><searchLink fieldCode="PB" term="%22Elsevier%22">Elsevier </searchLink><i>Country of Publication: </i>Netherlands <i>NLM ID: </i>9318488 <i>Publication Model: </i>Print-Electronic <i>Cited Medium: </i>Internet <i>ISSN: </i>1873-3778 (Electronic) <i>Linking ISSN: </i><searchLink fieldCode="IS" term="%2200219673%22">00219673 </searchLink><i>NLM ISO Abbreviation: </i>J Chromatogr A <i>Subsets: </i>MEDLINE
– Name: PublisherInfo
  Label: Imprint Name(s)
  Group: PubInfo
  Data: <i>Original Publication</i>: Amsterdam ; New York : Elsevier, 1993-
– Name: SubjectMESH
  Label: MeSH Terms
  Group: Su
  Data: <searchLink fieldCode="MM" term="%22Volatile+Organic+Compounds%22">Volatile Organic Compounds*</searchLink>/<searchLink fieldCode="MM" term="%22Volatile+Organic+Compounds+analysis%22">analysis</searchLink> <br /><searchLink fieldCode="MM" term="%22Gas+Chromatography-Mass+Spectrometry%22">Gas Chromatography-Mass Spectrometry*</searchLink>/<searchLink fieldCode="MM" term="%22Gas+Chromatography-Mass+Spectrometry+methods%22">methods</searchLink> <br /><searchLink fieldCode="MM" term="%22Forensic+Sciences%22">Forensic Sciences*</searchLink>/<searchLink fieldCode="MM" term="%22Forensic+Sciences+methods%22">methods</searchLink> <br /><searchLink fieldCode="MM" term="%22Ion+Mobility+Spectrometry%22">Ion Mobility Spectrometry*</searchLink>/<searchLink fieldCode="MM" term="%22Ion+Mobility+Spectrometry+methods%22">methods</searchLink> <br /><searchLink fieldCode="MM" term="%22Paper%22">Paper*</searchLink> <br /><searchLink fieldCode="MM" term="%22Machine+Learning%22">Machine Learning*</searchLink><br /><searchLink fieldCode="MH" term="%22Boosting+Machine+Learning+Algorithms%22">Boosting Machine Learning Algorithms</searchLink> ; <searchLink fieldCode="MH" term="%22Humans%22">Humans</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The determination of paper aging is crucial for forensic document examination, enabling the identification of temporal origin in forgery or fraud cases. Paper aging is difficult to distinguish due to the high compositional complexity, intricate temporal evolution mechanisms, and challenges in tracing diagnostic markers, limiting reliable chronological authentication. Herein, an interpretable generative end-to-end machine learning platform integrating gas chromatography-ion mobility spectrometry (GC-IMS) with machine learning was established for prediction of the temporal origin of forensic paper aging. Sixty-three differential volatiles were identified in paper materials via GC-IMS analysis. Chronological classification and regression machine learning models were constructed based the volatile fingerprints from papers aged 1 to 35 years. The CatBoost model achieved the best overall performance in both age-class identification and year prediction based on characteristic volatile profiles. SHAP (SHapley Additive exPlanations) model identified key markers as the most influential predictors, such as 3‑hydroxy-2-butanone-D, Hydroxyacetone, Cyclopentanone, and Isomenthone, enabling paper samples specific attribution of prediction outcomes. An end-to-end chronological prediction software tool was developed to automate data import, age prediction, and interpretation result generation, thereby improving its practicality and accessibility for forensic applications. Overall, this study provides a deployable analytical framework for interpretable chronological prediction of forensic paper aging and is expected to generate reliable results for judicial authentication.<br /> (Copyright © 2026 Elsevier B.V. All rights reserved.)
– Name: Abstract
  Label: Competing Interests
  Group: Ab
  Data: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
– Name: SubjectMinor
  Label: Contributed Indexing
  Group:
  Data: <i>Keywords: </i>End-to-end generative platform; Forensic paper aging; GC-IMS; Interpretable machine learning; Volatile organic compounds
– Name: NumberCAS
  Label: Substance Nomenclature
  Group: ID
  Data: 0 (Volatile Organic Compounds)
– Name: DateEntry
  Label: Entry Date(s)
  Group: Date
  Data: <i>Date Created: </i>20260602 <i>Date Completed: </i>20260629 <i>Latest Revision: </i>20260629
– Name: DateUpdate
  Label: Update Code
  Group: Date
  Data: 20260630
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1016/j.chroma.2026.467149
– Name: AN
  Label: PMID
  Group: ID
  Data: 42229280
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=cmedm&AN=42229280
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1016/j.chroma.2026.467149
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        StartPage: 467149
    Subjects:
      – SubjectFull: Boosting Machine Learning Algorithms
        Type: general
      – SubjectFull: Humans
        Type: general
      – SubjectFull: Volatile Organic Compounds analysis
        Type: general
      – SubjectFull: Gas Chromatography-Mass Spectrometry methods
        Type: general
      – SubjectFull: Forensic Sciences methods
        Type: general
      – SubjectFull: Ion Mobility Spectrometry methods
        Type: general
      – SubjectFull: Paper
        Type: general
      – SubjectFull: Machine Learning
        Type: general
    Titles:
      – TitleFull: Interpretable generative machine learning platform for chronological prediction of forensic paper aging via GC-IMS volatiles characterization in judicial authentication.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Lu W
      – PersonEntity:
          Name:
            NameFull: Kong L
      – PersonEntity:
          Name:
            NameFull: Zhang L
      – PersonEntity:
          Name:
            NameFull: Ma J
      – PersonEntity:
          Name:
            NameFull: Qi Y
      – PersonEntity:
          Name:
            NameFull: Lei M
      – PersonEntity:
          Name:
            NameFull: Li X
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 30
              M: 08
              Text: 2026 Aug 30
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-electronic
              Value: 1873-3778
          Numbering:
            – Type: volume
              Value: 1783
          Titles:
            – TitleFull: Journal of chromatography. A
              Type: main
ResultId 1