GaRSIVis: Improving the Predicting
of Self-Interruption during Reading using Gaze Data
Jan Pilzer, Shareen
Mahmud, Vanessa Putnam,
and Tamara Munzner. Proc. ETVIS 2018.
Paper
Talk
Software
Abstract
Gaze pattern data provides a promising opportunity to create a
predictive model of self-interruption during reading that could
support active interventions to keep a reader's attention at times
when self-interruptions are predicted to occur. We present two systems
designed to help analysts create and improve such a model. We present
GaRSIVis, (Gaze Reading Self-Interruption Visualizer), that integrates
a visualization front-end suitable for data cleansing and a prediction
back-end that can be run repeatedly as the input data is iteratively
improved. It allows analysts refining the predictive model to filter
out unwanted parts of the gaze data that should not be used in the
prediction. It relies on data gathered by GaRSILogger, which logs gaze
data and activity associated with interruptions during on-screen
reading. By integrating data cleansing and our prediction results in
our visualization, we enable analysts using GaRSIVis to come up with a
comprehensible way of understanding self-interruption from gaze
related features.
High-Resolution Figures
- Fig. 1. Overview of the interface showing a list of readings with their xation counts. Users can select a reading to clean the data or run a prediction on all readings to investigate the gaze pattern associated with a self interruption.
- Fig. 2 The Overview page (A) with the readings from which analysts can select a particular reading for cleaning. This reading is shown in the Data Cleansing page (B). Upon returning to the Overview page analysts can run a prediction on the cleaned data and view the Prediction Results page (C). From there analysts can choose to rechunk the data (D) and then rerun the prediction on the new chunk size. This workflow can be repeated.
- Fig. 3. Data Cleansing: brush and zoom to inspect data in detail and trim invalid portions.
- Fig. 4. Prediction Results: examine classified results from gaze related features for normal readings and readings before an interruption.
Last modified: Jun 16 2018.
