In the Software Campus Project SciBot, we investigated how users can be supported when seeking information through adaptation of a user interface. Concretely, we implemented and evaluated a method for estimating relevance of text passages based on eye movements and machine learning. We provide all related resources on this page.
gazeRE Dateset
We conducted a user study in which we asked users to rate the relevance of read texts with respect to a trigger question. We recorded the user’s gaze signal and their relevance ratings. Our dataset repository contains a set of scripts and routines to load, process, and analyse the recorded dataset. The goal is to estimate the user’s perceived relevance using machine learning with the gaze signal as input.
Interactive Data Exploration (Demo)
We present the Reading Model Assessment tool (ReMA), an interactive tool for assessing gaze-based relevance estimation models. Our tool allows experimenters to easily browse recorded trials, compare the model output to a ground truth, and visualize gaze-based features at the token- and paragraph-level that serve as model input. Our goal is to facilitate the understanding if the relation between eye movements and the human relevance estimation process, to understand the strengths and weaknesses of a model at hand.
Contact
Michael Barz
michael.barz@dfki.de
References
- Implicit Estimation of Paragraph Relevance from Eye Movements. In: Frontiers in Computer Science, 3 , pp. 13, 2022.
- Interactive Assessment Tool for Gaze-based Machine Learning Models in Information Retrieval. In: ACM SIGIR Conference on Human Information Interaction and Retrieval, Association for Computing Machinery, 2022.