Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Quantum Cognition-Inspired EEG-based Recommendation via Graph Neural Networks

Jinkun Han, Wei Li, Yingshu Li, Zhipeng Cai

TL;DR

The paper addresses real-time EEG-based item recommendation by introducing QUARK, a framework that fuses Quantum Cognition Theory with Graph Convolutional Networks to infer user thoughts from EEG signals. It segments EEG with sliding windows, encodes segments into quantum spaces, and uses continuity and interference graphs to capture temporal dependencies for ranking items via a neural graph learner. The model is trained with a three-term loss including BPR, orthogonality, and continuity terms, and evaluated on the MindBigData EEG dataset, where QUARK outperforms classical baselines and demonstrates interpretable representations of user feelings and style. These results suggest a feasible path toward EEG-driven, real-time personalization and point to future potential in cross-domain and multi-source brain-computer interface applications.

Abstract

Current recommendation systems recommend goods by considering users' historical behaviors, social relations, ratings, and other multi-modals. Although outdated user information presents the trends of a user's interests, no recommendation system can know the users' real-time thoughts indeed. With the development of brain-computer interfaces, it is time to explore next-generation recommenders that show users' real-time thoughts without delay. Electroencephalography (EEG) is a promising method of collecting brain signals because of its convenience and mobility. Currently, there is only few research on EEG-based recommendations due to the complexity of learning human brain activity. To explore the utility of EEG-based recommendation, we propose a novel neural network model, QUARK, combining Quantum Cognition Theory and Graph Convolutional Networks for accurate item recommendations. Compared with the state-of-the-art recommendation models, the superiority of QUARK is confirmed via extensive experiments.

Quantum Cognition-Inspired EEG-based Recommendation via Graph Neural Networks

TL;DR

The paper addresses real-time EEG-based item recommendation by introducing QUARK, a framework that fuses Quantum Cognition Theory with Graph Convolutional Networks to infer user thoughts from EEG signals. It segments EEG with sliding windows, encodes segments into quantum spaces, and uses continuity and interference graphs to capture temporal dependencies for ranking items via a neural graph learner. The model is trained with a three-term loss including BPR, orthogonality, and continuity terms, and evaluated on the MindBigData EEG dataset, where QUARK outperforms classical baselines and demonstrates interpretable representations of user feelings and style. These results suggest a feasible path toward EEG-driven, real-time personalization and point to future potential in cross-domain and multi-source brain-computer interface applications.

Abstract

Current recommendation systems recommend goods by considering users' historical behaviors, social relations, ratings, and other multi-modals. Although outdated user information presents the trends of a user's interests, no recommendation system can know the users' real-time thoughts indeed. With the development of brain-computer interfaces, it is time to explore next-generation recommenders that show users' real-time thoughts without delay. Electroencephalography (EEG) is a promising method of collecting brain signals because of its convenience and mobility. Currently, there is only few research on EEG-based recommendations due to the complexity of learning human brain activity. To explore the utility of EEG-based recommendation, we propose a novel neural network model, QUARK, combining Quantum Cognition Theory and Graph Convolutional Networks for accurate item recommendations. Compared with the state-of-the-art recommendation models, the superiority of QUARK is confirmed via extensive experiments.
Paper Structure (27 sections, 1 theorem, 21 equations, 6 figures, 4 tables, 1 algorithm)

This paper contains 27 sections, 1 theorem, 21 equations, 6 figures, 4 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Preliminaries
  4. Methodology
  5. Data Pre-processing & Sliding Windows
  6. Quantum Spaces & New States
  7. Matrices & Graph Convolutional Networks
  8. Continuity Adjacency Matrix
  9. Interference Adjacency Matrix
  10. Graph Convolutional Networks
  11. Model Training & Prediction
  12. Performance Evaluation
  13. Discussion & Conclusion
  14. Interference between Two Events
  15. Hyperparameter Analysis
  16. ...and 12 more sections

Key Result

theorem 1

If past event $\Theta_{m,1}$ influences the future event $\Theta_{m,2}$, the operator interpreting the event $\Theta_{m,1}$ occurring is defined as $o_{\Theta_{m,1}}=\sum {|b_j \rangle\langle b_j|}$ with selected basis vectors $\{b_j| selected \; j\}$, while the operator of $\Theta_{m,1}$ not occurr

Figures (6)

  • Figure 1: Understand the thoughts or needs, and recommend related items.
  • Figure 2: The framework of QUARK.
  • Figure 3: Representation visualization. (a) Similarity of raw EEG signals, (b) Similarity of representation learned by QUARK, (c) Similarity of representation learned by BPR, (d) Similarity of raw image data, and (e) Similarity of learned image representation.
  • Figure 4: The feeling and style of the recommended images compared with the original EEG images. (a) Content similarity. (b) Color similarity. (c) Structural similarity. (d) Synthesis score. (e) Synthesis score with precision@10.
  • Figure 5: Case studies.
  • ...and 1 more figures

Theorems & Definitions (1)

  • theorem 1