Skip to content

Read-only version: For security reasons, the Research Information System (FIS) has been temporarily replaced by this read-only version. All public content and full texts remain available; login, editing, submission and statistics are not available. Questions: fis@uni-bamberg.de

Home › Publications › Actor Identification in Discourse : A Challenge for LLMs?

Publication

Actor Identification in Discourse : A Challenge for LLMs?

Barić, Ana; Padó, Sebastian; Papay, Sean (2024): Actor Identification in Discourse : A Challenge for LLMs?, in: Michael Strube, Chloe Braud, Christian Hardmeier, u. a. (Hrsg.), Proceedings of the 5th Workshop on Computational Approaches to Discourse (CODI 2024), Association for Computational Linguistics, S. 64–70, doi: 10.18653/v1/2024.codi-1.6.

Details

Faculty/Chair
Faculty/Professorship
Author
Barić, Ana
Padó, Sebastian
Title of the compilation
Proceedings of the 5th Workshop on Computational Approaches to Discourse (CODI 2024)
Editors
Strube, Michael
Braud, Chloe
Hardmeier, Christian
Li, Junyi Jessy
Loaiciga, Sharid
Zeldes, Amir
Li, Chuyuan
Conference
5th Workshop on Computational Approaches to Discourse (CODI 2024), St Julian’s, Malta
Publisher Information
Association for Computational Linguistics
Year of publication
2024
Language
English
DOI
URL
Abstract

The identification of political actors who put forward claims in public debate is a crucial step in the construction of discourse networks, which are helpful to analyze societal debates. Actor identification is, however, rather challenging: Often, the locally mentioned speaker of a claim is only a pronoun (“He proposed that [claim]”), so recovering the canonical actor name requires discourse understanding. We compare a traditional pipeline of dedicated NLP components (similar to those applied to the related task of coreference) with a LLM, which appears a good match for this generation task. Evaluating on a corpus of German actors in newspaper reports, we find surprisingly that the LLM performs worse. Further analysis reveals that the LLM is very good at identifying the right reference, but struggles to generate the correct canonical form. This points to an underlying issue in LLMs with controlling generated output. Indeed, a hybrid model combining the LLM with a classifier to normalize its output substantially outperforms both initial models.

Keywords
Actor Identification
Peer Reviewed
yes
International Distribution
yes
Open Access Journal
yes
Type
Conferenceobject
URI
Activation date
2026-05-28

Permalink: https://fis.uni-bamberg.de/handle/uniba/115293 · Data as JSON · Cite: BibTeX · RIS

Data as of: 2026-11-15

Restricted read-only access: login, editing and submission are temporarily disabled.