# Towards Making a Dependency Parser See

arXiv: https://arxiv.org/abs/1909.01053 · HTML: https://mlsanity.com/paper/1909.01053/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Sep 2019.

## Built on by (most cited first, 5 of 5)

- [Synthesizing Human Gaze Feedback for Improved NLP Performance](https://mlsanity.com/paper/2302.05721/) (2023)
- [CogAlign: Learning to Align Textual Neural Representations to Cognitive Language Processing Signals](https://mlsanity.com/paper/2106.05544/) (2021)
- [TorontoCL at CMCL 2021 Shared Task: RoBERTa with Multi-Stage Fine-Tuning for Eye-Tracking Prediction](https://mlsanity.com/paper/2104.07244/) (2021)
- [Bridging between Cognitive Processing Signals and Linguistic Features via a Unified Attentional Network](https://mlsanity.com/paper/2112.08831/) (2021)
- [Eye-Tracking-while-Reading: A Living Survey of Datasets with Open Library Support](https://mlsanity.com/paper/2602.19598/) (2026)

## Abstract (arXiv)

We explore whether it is possible to leverage eye-tracking data in an RNN dependency parser (for English) when such information is only available during training, i.e., no aggregated or token-level gaze features are used at inference time. To do so, we train a multitask learning model that parses sentences as sequence labeling and leverages gaze features as auxiliary tasks. Our method also learns to train from disjoint datasets, i.e. it can be used to test whether already collected gaze features are useful to improve the performance on new non-gazed annotated treebanks. Accuracy gains are modest but positive, showing the feasibility of the approach. It can serve as a first step towards architectures that can better leverage eye-tracking data or other complementary information available only for training sentences, possibly leading to improvements in syntactic parsing.