RecGaze Dataset - Non-Public Version

This is the non-public RecGaze dataset from the paper: RecGaze: The First Eye Tracking and User Interaction Dataset for Carousel Interfaces

Link to open-acess paper: SIGIR 2025, Arxiv

Follow-up eye tracking analysis of user browsing behavior: IUI 2026

Please cite the following: 

@inproceedings{10.1145/3726302.3730301,
author = {de Leon-Martinez, Santiago and Kang, Jingwei and Moro, Robert and de Rijke, Maarten and Kveton, Branislav and Oosterhuis, Harrie and Bielikova, Maria},
title = {RecGaze: The First Eye Tracking and User Interaction Dataset for Carousel Interfaces},
year = {2025},
isbn = {9798400715921},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3726302.3730301},
doi = {10.1145/3726302.3730301},
booktitle = {Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval},
pages = {3702–3711},
numpages = {10},
keywords = {browsing behavior, carousel interfaces, eye tracking},
location = {Padua, Italy},
series = {SIGIR '25}
}

@inproceedings{10.1145/3742413.3789166,
author = {de Leon-Martinez, Santiago and Moro, Robert and Kveton, Branislav and Bielikova, Maria},
title = {Riding the Carousel: The First Extensive Eye Tracking Analysis of Browsing Behavior in Carousel Recommenders},
year = {2026},
isbn = {9798400719844},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3742413.3789166},
doi = {10.1145/3742413.3789166},
booktitle = {Proceedings of the 31st International Conference on Intelligent User Interfaces},
pages = {2120–2130},
numpages = {11},
keywords = {Carousel interfaces, Multi-list recommendations, Browsing behavior, Eye tracking},
location = {
},
series = {IUI '26}
}

Dataset Description

The RecGaze dataset is 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.

Non-Public Dataset download contains:

• Non-Public User Features Dataframe (non_public_user_features.csv) - Contains all the information gathered from the users during the pre-survey, post-survey, and post-selection screens (selection explanations). Non-public version adds:
  • Age
  • Gender
  • Answer to most helpful carousel topic/explanation question

• Non-public Feedback Dataframe (non_public_feedback_dataset.csv) - All the feedback (fixations, clicks, cursor movements) data gathered during the movie selection screens. Non-public version adds:
  • Raw Gaze data 
  • x,y pixel postions for fixation, cursor, clicks

• Poster Images (poster_images_resize.zip)- For convenience, we provide a zip folder including all the poster images of the movies used in the user study. They can also be accessed through the URL in the Item Features of the public version of the dataset.
• Screen Recordings - Includes recordings of every movie selection task for all users and screens. Note:
  •  Screen recordings include the loading (at start and after movie selection) of the webpage, for which all feedback data is already removed from the public and non-public summary feedback dataframe. 
  • The blue dot is a gaze indicator created by the post-processing of the recording software used by the user study.
  • Screen recordings [84GB] exceed the allowed storage amount for Zenodo. If you would like access to them then mention it in the email (see below). 

A more detailed description of all the files and their contents (along with supplementary material) can be found in the GitHub.

How to recieve access to the non-public version

1. Fill out the form below and in your request message state that you agree to the terms
2. Data sharing agreement complete (sections in red), sign, and email it to santiago.deleon@kinit.sk. If you would like access to the screen recordings then please mention it in the email.
3. Upon confirmation of 1 and 2, we will approve access to download from Zenodo