Publikation: Figurative Language Processing : A Linguistically Informed Feature Analysis of the Behavior of Language Models and Humans
Dateien
Datum
Autor:innen
Herausgeber:innen
ISSN der Zeitschrift
Electronic ISSN
ISBN
Bibliografische Daten
Verlag
Schriftenreihe
Auflagebezeichnung
DOI (zitierfähiger Link)
Internationale Patentnummer
Angaben zur Forschungsförderung
Projekt
Open Access-Veröffentlichung
Sammlungen
Core Facility der Universität Konstanz
Titel in einer weiteren Sprache
Publikationstyp
Publikationsstatus
Erschienen in
Zusammenfassung
Recent years have witnessed a growing interest in investigating what Transformer-based language models (TLMs) actually learn from the training data. This is especially relevant for complex tasks such as the understanding of non-literal meaning. In this work, we probe the performance of three black-box TLMs and two intrinsically transparent white-box models on figurative language classification of sarcasm, similes, idioms, and metaphors. We conduct two studies on the classification results to provide insights into the inner workings of such models. With our first analysis on feature importance, we identify crucial differences in model behavior. With our second analysis using an online experiment with human participants, we inspect different linguistic characteristics of the four figurative language types.
Zusammenfassung in einer weiteren Sprache
Fachgebiet (DDC)
Schlagwörter
Konferenz
Rezension
Zitieren
ISO 690
JANG, Hye Won, Qi YU, Diego FRASSINELLI, 2023. Figurative Language Processing : A Linguistically Informed Feature Analysis of the Behavior of Language Models and Humans. Findings of the Association for Computational Linguistics : ACL 2023. Toronto, Canada, 9. Juli 2023 - 14. Juli 2023. In: ROGERS, Anna, ed., Jordan BOYD-GRABER, ed., Naoaki OKAZAKI, ed.. Findings of the Association for Computational Linguistics : ACL 2023. Stroudsburg: Association for Computational Linguistics, 2023, pp. 9816-9832. Available under: doi: 10.18653/v1/2023.findings-acl.622BibTex
@inproceedings{Jang2023Figur-68828, year={2023}, doi={10.18653/v1/2023.findings-acl.622}, title={Figurative Language Processing : A Linguistically Informed Feature Analysis of the Behavior of Language Models and Humans}, publisher={Association for Computational Linguistics}, address={Stroudsburg}, booktitle={Findings of the Association for Computational Linguistics : ACL 2023}, pages={9816--9832}, editor={Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki}, author={Jang, Hye Won and Yu, Qi and Frassinelli, Diego} }
RDF
<rdf:RDF xmlns:dcterms="http://purl.org/dc/terms/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:bibo="http://purl.org/ontology/bibo/" xmlns:dspace="http://digital-repositories.org/ontologies/dspace/0.1.0#" xmlns:foaf="http://xmlns.com/foaf/0.1/" xmlns:void="http://rdfs.org/ns/void#" xmlns:xsd="http://www.w3.org/2001/XMLSchema#" > <rdf:Description rdf:about="https://kops.uni-konstanz.de/server/rdf/resource/123456789/68828"> <dcterms:title>Figurative Language Processing : A Linguistically Informed Feature Analysis of the Behavior of Language Models and Humans</dcterms:title> <dc:contributor>Yu, Qi</dc:contributor> <void:sparqlEndpoint rdf:resource="http://localhost/fuseki/dspace/sparql"/> <dcterms:available rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2023-12-20T08:38:10Z</dcterms:available> <dc:contributor>Frassinelli, Diego</dc:contributor> <foaf:homepage rdf:resource="http://localhost:8080/"/> <dc:creator>Frassinelli, Diego</dc:creator> <dc:creator>Jang, Hye Won</dc:creator> <dcterms:issued>2023</dcterms:issued> <dspace:isPartOfCollection rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/45"/> <dc:contributor>Jang, Hye Won</dc:contributor> <dc:language>eng</dc:language> <dc:date rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2023-12-20T08:38:10Z</dc:date> <dcterms:isPartOf rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/45"/> <dc:creator>Yu, Qi</dc:creator> <dcterms:abstract>Recent years have witnessed a growing interest in investigating what Transformer-based language models (TLMs) actually learn from the training data. This is especially relevant for complex tasks such as the understanding of non-literal meaning. In this work, we probe the performance of three black-box TLMs and two intrinsically transparent white-box models on figurative language classification of sarcasm, similes, idioms, and metaphors. We conduct two studies on the classification results to provide insights into the inner workings of such models. With our first analysis on feature importance, we identify crucial differences in model behavior. With our second analysis using an online experiment with human participants, we inspect different linguistic characteristics of the four figurative language types.</dcterms:abstract> <bibo:uri rdf:resource="https://kops.uni-konstanz.de/handle/123456789/68828"/> </rdf:Description> </rdf:RDF>