This repository contains the source code, data, and analyses behind the paper Using Gaze Data to Predict Multiword Expressions.
We have annotated the text used in GECO eye tracking corpus with regards to occurrences of Verb+Noun and Verb+Particle multiword expressions (MWEs). For details about the data, please refer to the information file in the data folder.
We used the structured prediction model CRF to label the sequences according to the classes defined in the tagged dataset.