Home About Subscribe Search Member Area

Humanist Discussion Group


< Back to Volume 32

Humanist Archives: March 5, 2019, 6:30 a.m. Humanist 32.512 - authorship attribution for automatically generated texts

                  Humanist Discussion Group, Vol. 32, No. 512.
            Department of Digital Humanities, King's College London
                   Hosted by King's Digital Lab
                       www.dhhumanist.org
                Submit to: humanist@dhhumanist.org




        Date: 2019-03-04 15:21:08+00:00
        From: John Lavagnino 
        Subject: Re: [Humanist] 32.503: authorship attribution for automatically generated texts?

There was recent work of exactly this kind that I thought was quite well
known---"Duplicate and fake publications in the scientific literature: how many
SCIgen papers in computer science?" by Cyril Labbé and Dominique Labbé from
2012, available in open-access form at

https://hal.archives-ouvertes.fr/hal-00641906v2

In brief: one group of computational linguists developed text-generation
software called SCIgen that writes computer-science papers; they're in the
right form, have grammatical sentences, but are nonsense.   Labbé and Labbé
built their own computational-linguistic system to detect SCIgen papers and
discovered there are quite a lot of them in published computer-science
conference proceedings.  There was subsequently a piece in Nature about the
retraction by publishers of some of these papers: see

https://www.nature.com/news/publishers-withdraw-more-than-120-gibberish-
papers-1.14763

Next time you see one of those stories about supposedly nonsensical publications
in humanities journals, this is the thing to remember: it's really computer
science that has the problem.

John


---
Dr John Lavagnino
Reader in Digital Humanities
Department of Digital Humanities and Department of English
King's College London
Strand
London WC2R 2LS
+44 20 7848 2453
www.lavagnino.org.uk




_______________________________________________
Unsubscribe at: http://dhhumanist.org/Restricted
List posts to: humanist@dhhumanist.org
List info and archives at at: http://dhhumanist.org
Listmember interface at: http://dhhumanist.org/Restricted/
Subscribe at: http://dhhumanist.org/membership_form.php


Editor: Willard McCarty (King's College London, U.K.; Western Sydney University, Australia)
Software designer: Malgosia Askanas (Mind-Crafts)

This site is maintained under a service level agreement by King's Digital Lab.