Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval

Leixin Zhang; Daniel Braun

doi:10.18653/v1/2024.argmining-1.17

Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval

Leixin Zhang, Daniel Braun

Industrial Engineering & Business Information Systems

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

1 Citation (Scopus)

130 Downloads (Pure)

Abstract

The paper describes our system for the Perspective Argument Retrieval Shared Task. The shared task consists of three scenarios in which relevant political arguments have to be retrieved based on queries (Scenario 1). In Scenario 2 explicit socio-cultural properties are provided and in Scenario 3 implicit socio-cultural properties within the arguments have to be used. We combined a Bi-Encoder and a Cross-Encoder to retrieve relevant arguments for each query. For the third scenario, we extracted linguistic features to predict socio-demographic labels as a separate task. However, the socio-demographic match task proved challenging due to the constraints of argument lengths and genres. The described system won both tracks (relevance and diversity) of the shared task.

Original language	English
Title of host publication	Proceedings of the 11th Workshop on Argument Mining (ArgMining 2024)
Editors	Yamen Ajjour, Roy Bar-Haim, Roxanne El Baff, Zhexiong Liu, Gabriella Skitalinskaya
Place of Publication	Bangkok, Thailand
Publisher	Association for Computational Linguistics (ACL)
Pages	164-168
Number of pages	5
ISBN (Electronic)	9798891761339
DOIs	https://doi.org/10.18653/v1/2024.argmining-1.17
Publication status	Published - 1 Aug 2024
Event	11th Workshop on Argument Mining, ArgMining 2024 - Bangkok, Thailand Duration: 15 Aug 2024 → 15 Aug 2024 Conference number: 11

Workshop

Workshop	11th Workshop on Argument Mining, ArgMining 2024
Abbreviated title	ArgMining 2024
Country/Territory	Thailand
City	Bangkok
Period	15/08/24 → 15/08/24

Access to Document

10.18653/v1/2024.argmining-1.17Licence: CC BY

2024.argmining-1.17Final published version, 162 KBLicence: CC BY

Cite this

Zhang, L., & Braun, D. (2024). Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval. In Y. Ajjour, R. Bar-Haim, R. El Baff, Z. Liu, & G. Skitalinskaya (Eds.), Proceedings of the 11th Workshop on Argument Mining (ArgMining 2024) (pp. 164-168). Association for Computational Linguistics (ACL). https://doi.org/10.18653/v1/2024.argmining-1.17

Zhang, Leixin ; Braun, Daniel. / Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval. Proceedings of the 11th Workshop on Argument Mining (ArgMining 2024). editor / Yamen Ajjour ; Roy Bar-Haim ; Roxanne El Baff ; Zhexiong Liu ; Gabriella Skitalinskaya. Bangkok, Thailand : Association for Computational Linguistics (ACL), 2024. pp. 164-168

@inproceedings{c322a77bec5045a192f97d1046da5ecc,

title = "Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval",

abstract = "The paper describes our system for the Perspective Argument Retrieval Shared Task. The shared task consists of three scenarios in which relevant political arguments have to be retrieved based on queries (Scenario 1). In Scenario 2 explicit socio-cultural properties are provided and in Scenario 3 implicit socio-cultural properties within the arguments have to be used. We combined a Bi-Encoder and a Cross-Encoder to retrieve relevant arguments for each query. For the third scenario, we extracted linguistic features to predict socio-demographic labels as a separate task. However, the socio-demographic match task proved challenging due to the constraints of argument lengths and genres. The described system won both tracks (relevance and diversity) of the shared task.",

author = "Leixin Zhang and Daniel Braun",

year = "2024",

month = aug,

day = "1",

doi = "10.18653/v1/2024.argmining-1.17",

language = "English",

pages = "164--168",

editor = "Yamen Ajjour and Roy Bar-Haim and {El Baff}, Roxanne and Zhexiong Liu and Gabriella Skitalinskaya",

booktitle = "Proceedings of the 11th Workshop on Argument Mining (ArgMining 2024)",

publisher = "Association for Computational Linguistics (ACL)",

address = "United States",

note = "11th Workshop on Argument Mining, ArgMining 2024, ArgMining 2024 ; Conference date: 15-08-2024 Through 15-08-2024",

}

Zhang, L & Braun, D 2024, Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval. in Y Ajjour, R Bar-Haim, R El Baff, Z Liu & G Skitalinskaya (eds), Proceedings of the 11th Workshop on Argument Mining (ArgMining 2024). Association for Computational Linguistics (ACL), Bangkok, Thailand, pp. 164-168, 11th Workshop on Argument Mining, ArgMining 2024, Bangkok, Thailand, 15/08/24. https://doi.org/10.18653/v1/2024.argmining-1.17

Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval. / Zhang, Leixin; Braun, Daniel.
Proceedings of the 11th Workshop on Argument Mining (ArgMining 2024). ed. / Yamen Ajjour; Roy Bar-Haim; Roxanne El Baff; Zhexiong Liu; Gabriella Skitalinskaya. Bangkok, Thailand: Association for Computational Linguistics (ACL), 2024. p. 164-168.

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

TY - GEN

T1 - Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval

AU - Zhang, Leixin

AU - Braun, Daniel

N1 - Conference code: 11

PY - 2024/8/1

Y1 - 2024/8/1

N2 - The paper describes our system for the Perspective Argument Retrieval Shared Task. The shared task consists of three scenarios in which relevant political arguments have to be retrieved based on queries (Scenario 1). In Scenario 2 explicit socio-cultural properties are provided and in Scenario 3 implicit socio-cultural properties within the arguments have to be used. We combined a Bi-Encoder and a Cross-Encoder to retrieve relevant arguments for each query. For the third scenario, we extracted linguistic features to predict socio-demographic labels as a separate task. However, the socio-demographic match task proved challenging due to the constraints of argument lengths and genres. The described system won both tracks (relevance and diversity) of the shared task.

AB - The paper describes our system for the Perspective Argument Retrieval Shared Task. The shared task consists of three scenarios in which relevant political arguments have to be retrieved based on queries (Scenario 1). In Scenario 2 explicit socio-cultural properties are provided and in Scenario 3 implicit socio-cultural properties within the arguments have to be used. We combined a Bi-Encoder and a Cross-Encoder to retrieve relevant arguments for each query. For the third scenario, we extracted linguistic features to predict socio-demographic labels as a separate task. However, the socio-demographic match task proved challenging due to the constraints of argument lengths and genres. The described system won both tracks (relevance and diversity) of the shared task.

UR - http://www.scopus.com/inward/record.url?scp=85204927725&partnerID=8YFLogxK

U2 - 10.18653/v1/2024.argmining-1.17

DO - 10.18653/v1/2024.argmining-1.17

M3 - Conference contribution

SP - 164

EP - 168

BT - Proceedings of the 11th Workshop on Argument Mining (ArgMining 2024)

A2 - Ajjour, Yamen

A2 - Bar-Haim, Roy

A2 - El Baff, Roxanne

A2 - Liu, Zhexiong

A2 - Skitalinskaya, Gabriella

PB - Association for Computational Linguistics (ACL)

CY - Bangkok, Thailand

T2 - 11th Workshop on Argument Mining, ArgMining 2024

Y2 - 15 August 2024 through 15 August 2024

ER -

Zhang L, Braun D. Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval. In Ajjour Y, Bar-Haim R, El Baff R, Liu Z, Skitalinskaya G, editors, Proceedings of the 11th Workshop on Argument Mining (ArgMining 2024). Bangkok, Thailand: Association for Computational Linguistics (ACL). 2024. p. 164-168 doi: 10.18653/v1/2024.argmining-1.17

Twente-BMS-NLP at PerspectiveArg 2024: Combining Bi-Encoder and Cross-Encoder for Argument Retrieval

Abstract

Workshop

Access to Document

Other files and links

Fingerprint

Cite this