This paper describes the submission of the team KonTra to the CMCL 2021 Shared Task on eye-tracking prediction. Our system combines the embeddings extracted from a finetuned BERT model with surface, linguistic and behavioral features, resulting in an average mean absolute error of 4.22 across all 5 eyetracking measures. We show that word length and features representing the expectedness of a word are consistently the strongest predictors across all 5 eye-tracking measures.
CITATION STYLE
Yu, Q., Kalouli, A. L., & Frassinelli, D. (2021). KonTra at CMCL 2021 Shared Task: Predicting Eye Movements by Combining BERT with Surface, Linguistic and Behavioral Information. In CMCL 2021 - Workshop on Cognitive Modeling and Computational Linguistics, Proceedings (pp. 120–124). Association for Computational Linguistics (ACL). https://doi.org/10.18653/v1/2021.cmcl-1.15
Mendeley helps you to discover research relevant for your work.