Measuring the Impact of (Psycho-)Linguistic and Readability Features and Their Spill Over Effects on the Prediction of Eye Movement Patterns

TL;DR

This study investigates how various linguistic and readability features, along with transformer model architectures, influence the prediction of eye movement patterns during natural reading, revealing the importance of feature selection and model design.

Contribution

It systematically evaluates the impact of diverse text features and transformer architectures on eye-tracking prediction, highlighting their combined effects.

Findings

Both features and model architecture significantly affect prediction accuracy.

Certain linguistic features are more influential in predicting eye movements.

Transformer models' performance varies with feature inclusion.

Abstract

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.