HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features

Lavinia Salicchi; Rong Xiang; Yu-yin Hsu

doi:10.18653/v1/2022.cmcl-1.13

HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features

Lavinia Salicchi, Rong Xiang, Yu-yin Hsu

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

3 Citations (Scopus)

Abstract

Eye movement data are used in psycholinguistic studies to infer information regarding cognitive processes during reading. In this paper, we describe our proposed method for the Shared Task of Cognitive Modeling and Computational Linguistics (CMCL) 2022 - Subtask 1, which involves data from multiple datasets on 6 languages. We compared different regression models using features of the target word and its previous word, and target word surprisal as regression features. Our final system, using a gradient boosting regressor, achieved the lowest mean absolute error (MAE), resulting in the best system of the competition.

Original language	English
Title of host publication	Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics
Editors	Emmanuele Chersoni, Nora Hollenstein, Cassandra Jacobs, Yohei Oseki, Laurent Prévot, Enrico Santus
Pages	114–120
DOIs	https://doi.org/10.18653/v1/2022.cmcl-1.13
Publication status	Published - Mar 2022
Event	Workshop on Cognitive Modeling and Computational Linguistics (CMCL) 2022 - Dublin, Ireland Duration: 26 Apr 2022 → 26 Apr 2022 https://cmclorg.github.io/

Competition

Competition	Workshop on Cognitive Modeling and Computational Linguistics (CMCL) 2022
Country/Territory	Ireland
City	Dublin
Period	26/04/22 → 26/04/22
Internet address	https://cmclorg.github.io/

Keywords

gradient boosting
eyetracking
prediction
linguistic features
crosslingual

More information

10.18653/v1/2022.cmcl-1.13

http://hdl.handle.net/10397/95826

Cite this

Salicchi, L., Xiang, R., & Hsu, Y. (2022). HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features. In E. Chersoni, N. Hollenstein, C. Jacobs, Y. Oseki, L. Prévot, & E. Santus (Eds.), Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics (pp. 114–120) https://doi.org/10.18653/v1/2022.cmcl-1.13

Salicchi, Lavinia ; Xiang, Rong ; Hsu, Yu-yin. / HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features. Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics. editor / Emmanuele Chersoni ; Nora Hollenstein ; Cassandra Jacobs ; Yohei Oseki ; Laurent Prévot ; Enrico Santus. 2022. pp. 114–120

@inproceedings{6d1f8bd24e204e93b0b8a27b6f8157a2,

title = "HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features",

abstract = "Eye movement data are used in psycholinguistic studies to infer information regarding cognitive processes during reading. In this paper, we describe our proposed method for the Shared Task of Cognitive Modeling and Computational Linguistics (CMCL) 2022 - Subtask 1, which involves data from multiple datasets on 6 languages. We compared different regression models using features of the target word and its previous word, and target word surprisal as regression features. Our final system, using a gradient boosting regressor, achieved the lowest mean absolute error (MAE), resulting in the best system of the competition.",

keywords = "gradient boosting, eyetracking, prediction, linguistic features, crosslingual",

author = "Lavinia Salicchi and Rong Xiang and Yu-yin Hsu",

year = "2022",

month = mar,

doi = "10.18653/v1/2022.cmcl-1.13",

language = "English",

isbn = "978-1-955917-29-2",

pages = "114–120",

editor = "Emmanuele Chersoni and Nora Hollenstein and Cassandra Jacobs and Yohei Oseki and Laurent Pr{\'e}vot and Enrico Santus",

booktitle = "Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics",

note = "Workshop on Cognitive Modeling and Computational Linguistics (CMCL) 2022 ; Conference date: 26-04-2022 Through 26-04-2022",

url = "https://cmclorg.github.io/",

}

Salicchi, L, Xiang, R & Hsu, Y 2022, HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features. in E Chersoni, N Hollenstein, C Jacobs, Y Oseki, L Prévot & E Santus (eds), Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics. pp. 114–120, Workshop on Cognitive Modeling and Computational Linguistics (CMCL) 2022, Dublin, Ireland, 26/04/22. https://doi.org/10.18653/v1/2022.cmcl-1.13

HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features. / Salicchi, Lavinia; Xiang, Rong; Hsu, Yu-yin.
Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics. ed. / Emmanuele Chersoni; Nora Hollenstein; Cassandra Jacobs; Yohei Oseki; Laurent Prévot; Enrico Santus. 2022. p. 114–120.

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features

AU - Salicchi, Lavinia

AU - Xiang, Rong

AU - Hsu, Yu-yin

PY - 2022/3

Y1 - 2022/3

N2 - Eye movement data are used in psycholinguistic studies to infer information regarding cognitive processes during reading. In this paper, we describe our proposed method for the Shared Task of Cognitive Modeling and Computational Linguistics (CMCL) 2022 - Subtask 1, which involves data from multiple datasets on 6 languages. We compared different regression models using features of the target word and its previous word, and target word surprisal as regression features. Our final system, using a gradient boosting regressor, achieved the lowest mean absolute error (MAE), resulting in the best system of the competition.

AB - Eye movement data are used in psycholinguistic studies to infer information regarding cognitive processes during reading. In this paper, we describe our proposed method for the Shared Task of Cognitive Modeling and Computational Linguistics (CMCL) 2022 - Subtask 1, which involves data from multiple datasets on 6 languages. We compared different regression models using features of the target word and its previous word, and target word surprisal as regression features. Our final system, using a gradient boosting regressor, achieved the lowest mean absolute error (MAE), resulting in the best system of the competition.

KW - gradient boosting

KW - eyetracking

KW - prediction

KW - linguistic features

KW - crosslingual

U2 - 10.18653/v1/2022.cmcl-1.13

DO - 10.18653/v1/2022.cmcl-1.13

M3 - Conference article published in proceeding or book

SN - 978-1-955917-29-2

SP - 114

EP - 120

BT - Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics

A2 - Chersoni, Emmanuele

A2 - Hollenstein, Nora

A2 - Jacobs, Cassandra

A2 - Oseki, Yohei

A2 - Prévot, Laurent

A2 - Santus, Enrico

T2 - Workshop on Cognitive Modeling and Computational Linguistics (CMCL) 2022

Y2 - 26 April 2022 through 26 April 2022

ER -

HkAmsters at CMCL 2022 Shared Task: Predicting Eye-Tracking Data from a Gradient Boosting Framework with Linguistic Features

Abstract

Competition

Keywords

More information

Fingerprint

Cite this