RecGaze: The First Eye Tracking and User Interaction Dataset for Carousel Interfaces

TL;DR

Carousels are widespread but understudied in terms of user gaze and interaction. This paper introduces RecGaze, the first publicly available dataset that couples eye-tracking with interaction signals for carousel interfaces, including gaze, cursor, clicks, and selection explanations from 87 participants across 3 movie-selection tasks. The dataset comprises both public AOI-labeled data and a non-public raw data resource, organized into multiple feature-rich CSVs, along with preprocessing guidelines and use-case scenarios such as click modeling and offline evaluation. Initial gaze visualizations reveal a golden-triangle or F-pattern browsing behavior in carousels, supporting the need for gaze-informed recommender design and carousel-aware evaluation. Overall, RecGaze enables more empirical understanding of carousel interactions and supports development of gaze-based recommender systems and improved interface designs.

Abstract

Carousel interfaces are widely used in e-commerce and streaming services, but little research has been devoted to them. Previous studies of interfaces for presenting search and recommendation results have focused on single ranked lists, but it appears their results cannot be extrapolated to carousels due to the added complexity. Eye tracking is a highly informative approach to understanding how users click, yet there are no eye tracking studies concerning carousels. There are very few interaction datasets on recommenders with carousel interfaces and none that contain gaze data. We introduce the RecGaze dataset: the first comprehensive feedback dataset on carousels that includes eye tracking results, clicks, cursor movements, and selection explanations. The dataset comprises of interactions from 3 movie selection tasks with 40 different carousel interfaces per user. In total, 87 users and 3,477 interactions are logged. In addition to the dataset, its description and possible use cases, we provide results of a survey on carousel design and the first analysis of gaze data on carousels, which reveals a golden triangle or F-pattern browsing behavior. Our work seeks to advance the field of carousel interfaces by providing the first dataset with eye tracking results on carousels. In this manner, we provide and encourage an empirical understanding of interactions with carousel interfaces, for building better recommender systems through gaze information, and also encourage the development of gaze-based recommenders.

Figures (5)

• Figure 1: A capture of a free-browsing screen at initial presentation with the first 3 carousels shown and mouse-over details of a movie in the top row. Movies are ranked (left to right) by popularity based on IMDB number of votes. Movie posters and mouse-over details were shown (removed here due to copyright). See GitHub below for a sample screen recording gif.
• Figure 2: How familiar users are with the selected movie.
• Figure 3: Why users select a movie (multiple responses).
• Figure 4: Aggregate fixation heat map for every user on all 30 free-browsing screens with duration weighting shown on a to-scale background stimuli of the movie screens (with movie posters shown as blue and genre text as white boxes). It includes horizontal displacement, so initial 5 movies can be distinguished from second and third set from swiping.
• Figure 5: Visualizations of fixations, clicks and mouse cursor movements for one selected user and one free-browsing screen.