DedupEndNote: Changelog

1.1.2 (2025-12-30)

  • Several small changes in preprocessing and comparison of pages and titles further reducing the False Positives and False Negatives
  • Upgrade to Spring Boot 3.5.9
  • More refactoring of the code for normalization and comparison

1.1.1 (2025-11-08)

  • Enrichment of the pages field of deduplicated records has been corrected: complex pages ("1252-3; author reply 1253") were truncated to the first part
  • Refactoring of the code for normalization and comparison
  • Upgrade to Spring Boot 3.5.7

1.1.0 (2025-10-08)

  • Some changes in the comparisons:
    • Step 2: If at least one of the two publications has more than 2 pages, the DOIs are compared before the starting pages are compared
    • Step 3: Only the first 40 authors are compared
    • Step 4: Changes in splitting titles with subtitles. Most common languages at the end of title are removed. If one or both titles are empty, the answer is "YES"
    • Step 5: If both records have a ISBN, they are compared. If they are not the same, step 5 gets the answer "NO"
  • Records originally from clinicaltrials.gov but imported from clinicaltrials.gov, Cochrane Library and EMBASE are deduplicated much better and the deduplicated output is standardized. See Details: Special cases: ClinicalTrials.gov
  • Better handling of conference proceedings and books. Triggered by a new test file with a query on AI. The changed comparison by ISBN (see above) was triggered by this.
  • Errata and comments are treated in the same way as replies ("step 4: compare titles" is skipped; and in "enrich the deduplicated records" the longest duplicate title is used)
  • Retractions / retracted publications: in the preprocessing all parts of the title referring to the retraction are removed, and in "enrich the deduplicated records" the longest duplicate title is used). There is no special treatment of retraction notices
  • Comparison of export and deduplication of same publications in EndNote and Zotero
  • The text on the front page is simplified / shortened (with the help of Copilot). The full details are merged and updated in one file, appropriately called "Details, lots of details ...".
  • Upgrade to Spring Boot 3.5.6, Bootstrap 5.3.8

1.0.2 (2025-08-01)

  • Changed 5th comparison: this comparison is skipped if both publications have the same DOI
  • Enrich deduplicated records: missing journal titles added (esp. "clinicaltrials.gov", and "Social Science Research Network" from Embase.com)
  • Upgrade to Java 21, Spring Boot 3.5.4
*: A next release will try to handle ClinicalTrials.gov records (from ClinicalTrials.gov, Cochrane, Embase (OVID en Embase.com)) better

Older versions

  • 2024-01-14: 1.0.1: BUG: Doesn't handle author last name starting with a hyphen
  • 2023-09-02: ISBN-10 and ISBN-13 are compared correctly: 9024274214 = 9789024274215
  • 2023-09-02: BUG: Publication years of form "1-1-1989" (instead of "1989") are parsed as years
  • 2023-08-21: If a title with ": " has at least 50 characters before the ": ", that first part is also used as a title variant
  • 2023-08-21: BUG: Result file for deduplication of two files in Mark mode was not always written
  • 2023-08-18: "How to cite" added (CITATION.cff)
  • 2023-08-12: Files without IDs (record numbers) for each record (e.g. Zotero), get an ID assigned by DedupEndNote
  • 2023-08-12: Deduplication of Conference proceedings from Web of Science was very poor (too many false positives). This problem has been solved