Evaluating the Robustness of Embedding-based Topic Models to OCR Noise

Elaine Zosa, Mark Granroth-Wilding, Stephen Mutuvi, Antoine Doucet

Forskningsoutput: Kapitel i bok/rapport/konferenshandlingKonferensbidragVetenskapligPeer review

Sammanfattning

Unsupervised topic models such as Latent Dirichlet Allocation (LDA) are popular tools to analyse digitised corpora. However, the performance of these tools have been shown to degrade with OCR noise. Topic models that incorporate word embeddings during inference have been proposed to address the limitations of LDA, but these models have not seen much use in historical text analysis. In this paper we explore the impact of OCR noise on two embedding-based models, Gaussian LDA and the Embedded Topic Model (ETM) and compare their performance to LDA. Our results show that these models, especially ETM, are slightly more resilient than LDA in the presence of noise in terms of topic quality and classification accuracy.
Originalspråkengelska
Titel på värdpublikationTowards Open and Trustworthy Digital Societies. ICADL 2021
RedaktörerHao-Ren Ke, Chei Sian Lee, Kazunari Sugiyama
Antal sidor9
UtgivningsortCham
FörlagSpringer
Utgivningsdatum30 nov. 2021
Sidor392-400
ISBN (tryckt)978-3-030-91668-8
ISBN (elektroniskt)978-3-030-91669-5
DOI
StatusPublicerad - 30 nov. 2021
MoE-publikationstypA4 Artikel i en konferenspublikation
EvenemangInternational Conference on Asia-Pacific Digital Libraries - online
Varaktighet: 1 dec. 20213 dec. 2021
Konferensnummer: 23
https://icadl.net/icadl2021/

Publikationsserier

NamnLecture Notes in Computer Science
Volym13133
ISSN (tryckt)0302-9743
ISSN (elektroniskt)1611-3349

Vetenskapsgrenar

  • 113 Data- och informationsvetenskap

Citera det här