How About Time? Probing a Multilingual Language Model for Temporal Relations

Tommaso Caselli; Irene Dini; Felice Dell'Orletta

How About Time? Probing a Multilingual Language Model for Temporal Relations

Tommaso Caselli, Irene Dini, Felice Dell'Orletta

Computational Linguistics (CL)

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

1 Citation (Scopus)

52 Downloads (Pure)

Abstract

This paper presents a comprehensive set of probing experiments using a multilingual language model, XLM-R, for temporal relation classification between events in four languages. Results show an advantage of contextualized embeddings over static ones and a detrimen- tal role of sentence level embeddings. While obtaining competitive results against state-of-the-art systems, our probes indicate a lack of suitable encoded information to properly address this task.
pdf bib abs

Original language	English
Title of host publication	Proceedings of the 29th International Conference on Computational Linguistics
Place of Publication	Gyeongju, Republic of Korea
Publisher	International Committee on Computational Linguistics (ICCL)
Pages	3197-3209
Number of pages	13
Publication status	Published - 2022
Event	29th International Conference on Computational Linguistics - Gyeongju, Korea, Republic of Duration: 12-Oct-2022 → 17-Oct-2022

Conference

Conference	29th International Conference on Computational Linguistics
Abbreviated title	COLING 2022
Country/Territory	Korea, Republic of
City	Gyeongju
Period	12/10/2022 → 17/10/2022

Keywords

probing
temporal relation
event

Access to Document

How about Time? Probing a Multilingual Language Model for Temporal RelationsFinal publisher's version, 397 KBLicence: CC BY

Handle.net

Persistent link

Cite this

@inproceedings{9288efc1356a4e8ab10a008cc8ed5603,

title = "How About Time? Probing a Multilingual Language Model for Temporal Relations",

abstract = "This paper presents a comprehensive set of probing experiments using a multilingual language model, XLM-R, for temporal relation classification between events in four languages. Results show an advantage of contextualized embeddings over static ones and a detrimen- tal role of sentence level embeddings. While obtaining competitive results against state-of-the-art systems, our probes indicate a lack of suitable encoded information to properly address this task.pdf bib abs",

keywords = "probing, temporal relation, event",

author = "Tommaso Caselli and Irene Dini and Felice Dell'Orletta",

year = "2022",

language = "English",

pages = "3197--3209",

booktitle = "Proceedings of the 29th International Conference on Computational Linguistics",

publisher = "International Committee on Computational Linguistics (ICCL)",

note = "29th International Conference on Computational Linguistics, COLING 2022 ; Conference date: 12-10-2022 Through 17-10-2022",

}

Caselli, T, Dini, I & Dell'Orletta, F 2022, How About Time? Probing a Multilingual Language Model for Temporal Relations. in Proceedings of the 29th International Conference on Computational Linguistics. International Committee on Computational Linguistics (ICCL), Gyeongju, Republic of Korea, pp. 3197-3209, 29th International Conference on Computational Linguistics, Gyeongju, Korea, Republic of, 12/10/2022.

How About Time? Probing a Multilingual Language Model for Temporal Relations. / Caselli, Tommaso; Dini, Irene; Dell'Orletta, Felice.
Proceedings of the 29th International Conference on Computational Linguistics. Gyeongju, Republic of Korea: International Committee on Computational Linguistics (ICCL), 2022. p. 3197-3209.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - How About Time? Probing a Multilingual Language Model for Temporal Relations

AU - Caselli, Tommaso

AU - Dini, Irene

AU - Dell'Orletta, Felice

PY - 2022

Y1 - 2022

N2 - This paper presents a comprehensive set of probing experiments using a multilingual language model, XLM-R, for temporal relation classification between events in four languages. Results show an advantage of contextualized embeddings over static ones and a detrimen- tal role of sentence level embeddings. While obtaining competitive results against state-of-the-art systems, our probes indicate a lack of suitable encoded information to properly address this task.pdf bib abs

AB - This paper presents a comprehensive set of probing experiments using a multilingual language model, XLM-R, for temporal relation classification between events in four languages. Results show an advantage of contextualized embeddings over static ones and a detrimen- tal role of sentence level embeddings. While obtaining competitive results against state-of-the-art systems, our probes indicate a lack of suitable encoded information to properly address this task.pdf bib abs

KW - probing

KW - temporal relation

KW - event

UR - https://aclanthology.org/2022.coling-1.283/

M3 - Conference contribution

SP - 3197

EP - 3209

BT - Proceedings of the 29th International Conference on Computational Linguistics

PB - International Committee on Computational Linguistics (ICCL)

CY - Gyeongju, Republic of Korea

T2 - 29th International Conference on Computational Linguistics

Y2 - 12 October 2022 through 17 October 2022

ER -

How About Time? Probing a Multilingual Language Model for Temporal Relations

Abstract

Conference

Keywords

Access to Document

Handle.net

Other files and links

Fingerprint

Outstanding Paper - COLING 2022

Cite this

How About Time? Probing a Multilingual Language Model for Temporal Relations

Abstract

Conference

Keywords

Access to Document

Handle.net

Other files and links

Fingerprint

Prizes

Outstanding Paper - COLING 2022

Cite this