Measuring the Impact of (Psycho-)Linguistic and Readability Features and Their Spill Over Effects on the Prediction of Eye Movement Patterns
| Autor: | Wiechmann, D., Qiao, Y., Kerz, E., Mattern, J., Muresan, S., Nakov, P., Villavicencio, A. |
|---|---|
| Přispěvatelé: | ILLC (FGw) |
| Jazyk: | angličtina |
| Rok vydání: | 2022 |
| Předmět: | |
| Zdroj: | Stroudsburg, PA : Association for Computational Linguistics (ACL) 5267-5290 (2022). doi:10.18653/v1/2022.acl-long.362 The 60th Annual Meeting of the Association for Computational Linguistics-proceedings of the conference.-Vol. 1: Long papers The 60th Annual Meeting of the Association for Computational Linguistics-proceedings of the conference.-Vol. 1: Long papers60. Annual Meeting of the Association-for-Computational-Linguistics, ACL, Dublin, Ireland, 2022-05-22-2022-05-27 The 60th Annual Meeting of the Association for Computational Linguistics: ACL 2022 : proceedings of the conference : May 22-27, 2022, 1 |
| DOI: | 10.18154/rwth-2022-09014 |
| Popis: | There is a growing interest in the combined use of NLP and machine learning methods to predict gaze patterns during naturalistic reading. While promising results have been obtained through the use of transformer-based language models, little work has been undertaken to relate the performance of such models to general text characteristics. In this paper we report on experiments with two eye-tracking corpora of naturalistic reading and two language models (BERT and GPT-2). In all experiments, we test effects of a broad spectrum of features for predicting human reading behavior that fall into five categories (syntactic complexity, lexical richness, register-based multiword combinations, readability and psycholinguistic word properties). Our experiments show that both the features included and the architecture of the transformer-based language models play a role in predicting multiple eye-tracking measures during naturalistic reading. We also report the results of experiments aimed at determining the relative importance of features from different groups using SP-LIME. Comment: accepted at ACL 2022 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|