Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

EEG-SVRec: An EEG Dataset with User Multidimensional Affective Engagement Labels in Short Video Recommendation

Shaorun Zhang, Zhiyu He, Ziyi Ye, Peijie Sun, Qingyao Ai, Min Zhang, Yiqun Liu

TL;DR

EEG-SVRec introduces the first EEG-enabled dataset for short-video recommendation, combining EEG and ECG signals with multidimensional affective engagement scores (MAES) and user behavior in a realistic browsing setting. The dataset covers 30 participants with 3,657 interactions over 2,636 videos, and provides preprocessing pipelines, DE-based EEG features across five bands for 62 channels, and video-attribute features from both visual and audio modalities. Experiments show that incorporating EEG signals into rating-prediction and recommendation models yields significant performance gains, demonstrating the value of brain-based signals for understanding user cognition and affect in recommender systems. This work offers a new resource for human-centric evaluation and EEG-informed short-video recommendation research, with potential impacts on personalized, accessible, and more nuanced content discovery.

Abstract

In recent years, short video platforms have gained widespread popularity, making the quality of video recommendations crucial for retaining users. Existing recommendation systems primarily rely on behavioral data, which faces limitations when inferring user preferences due to issues such as data sparsity and noise from accidental interactions or personal habits. To address these challenges and provide a more comprehensive understanding of user affective experience and cognitive activity, we propose EEG-SVRec, the first EEG dataset with User Multidimensional Affective Engagement Labels in Short Video Recommendation. The study involves 30 participants and collects 3,657 interactions, offering a rich dataset that can be used for a deeper exploration of user preference and cognitive activity. By incorporating selfassessment techniques and real-time, low-cost EEG signals, we offer a more detailed understanding user affective experiences (valence, arousal, immersion, interest, visual and auditory) and the cognitive mechanisms behind their behavior. We establish benchmarks for rating prediction by the recommendation algorithm, showing significant improvement with the inclusion of EEG signals. Furthermore, we demonstrate the potential of this dataset in gaining insights into the affective experience and cognitive activity behind user behaviors in recommender systems. This work presents a novel perspective for enhancing short video recommendation by leveraging the rich information contained in EEG signals and multidimensional affective engagement scores, paving the way for future research in short video recommendation systems.

EEG-SVRec: An EEG Dataset with User Multidimensional Affective Engagement Labels in Short Video Recommendation

TL;DR

EEG-SVRec introduces the first EEG-enabled dataset for short-video recommendation, combining EEG and ECG signals with multidimensional affective engagement scores (MAES) and user behavior in a realistic browsing setting. The dataset covers 30 participants with 3,657 interactions over 2,636 videos, and provides preprocessing pipelines, DE-based EEG features across five bands for 62 channels, and video-attribute features from both visual and audio modalities. Experiments show that incorporating EEG signals into rating-prediction and recommendation models yields significant performance gains, demonstrating the value of brain-based signals for understanding user cognition and affect in recommender systems. This work offers a new resource for human-centric evaluation and EEG-informed short-video recommendation research, with potential impacts on personalized, accessible, and more nuanced content discovery.

Abstract

In recent years, short video platforms have gained widespread popularity, making the quality of video recommendations crucial for retaining users. Existing recommendation systems primarily rely on behavioral data, which faces limitations when inferring user preferences due to issues such as data sparsity and noise from accidental interactions or personal habits. To address these challenges and provide a more comprehensive understanding of user affective experience and cognitive activity, we propose EEG-SVRec, the first EEG dataset with User Multidimensional Affective Engagement Labels in Short Video Recommendation. The study involves 30 participants and collects 3,657 interactions, offering a rich dataset that can be used for a deeper exploration of user preference and cognitive activity. By incorporating selfassessment techniques and real-time, low-cost EEG signals, we offer a more detailed understanding user affective experiences (valence, arousal, immersion, interest, visual and auditory) and the cognitive mechanisms behind their behavior. We establish benchmarks for rating prediction by the recommendation algorithm, showing significant improvement with the inclusion of EEG signals. Furthermore, we demonstrate the potential of this dataset in gaining insights into the affective experience and cognitive activity behind user behaviors in recommender systems. This work presents a novel perspective for enhancing short video recommendation by leveraging the rich information contained in EEG signals and multidimensional affective engagement scores, paving the way for future research in short video recommendation systems.
Paper Structure (23 sections, 1 equation, 6 figures, 3 tables)

This paper contains 23 sections, 1 equation, 6 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Datasets
  3. Dataset in Short Video Recommendation
  4. EEG Dataset in Affective Computing
  5. Dataset Construction
  6. Ethical and Privacy
  7. Participants
  8. Video Stimuli Material
  9. Apparatus
  10. Experiemental Procedure
  11. Dataset Description
  12. EEG statistics and preprocessing
  13. User Behavior log and self-assessment of MAES
  14. Characteristic of Short Videos
  15. Example applications
  16. ...and 8 more sections

Figures (6)

  • Figure 1: EEG and ECG data acquisition setup: (a) A participant wears an EEG cap while watching short videos in a laboratory setting (Image display has been approved). (b) International 10-20 electrode placement standard for EEG.
  • Figure 2: The overall procedure of the lab study for data collection.
  • Figure 3: (a) Proportion of likes for short videos: overall and across three session modes (personalized, randomized, and mixed). (b) View percentage distribution across different session modes (View percentage is the viewing duration divided by the video duration. 1.0 represents viewing the video once.)
  • Figure 4: The distribution for six MAES (valence, arousal, immersion, interest, visual and auditory).
  • Figure 5: Heatmap presents the correlations of behavior (liking, and view percentage) and MAES (valence, arousal, immersion, interest, visual, and auditory).
  • ...and 1 more figures