QDA: Quantum-inspired Dynamic Adaptive Framework for Textual Sentiment Analysis

This is the official repository for the paper: "QDA: An End-to-End Quantum-inspired Dynamic Adaptive Framework for Textual Sentiment Analysis".

News / Updates

• [2025-01-27]: The paper has been completed.
• [2025-01-27]: Datasets have been uploaded to dataset/.

Abstract

Modeling with quantum theory shows promise in NLP, yet most rely on fixed interference terms and lack dynamic adaptive mechanisms, limiting their ability to handle varying sentence-level semantic ambiguity.

To this end, we propose QDA, an end-to-end Quantum-inspired Dynamic Adaptive framework for textual sentiment analysis. QDA consists of three designed modules that embed text as quantum states, adaptively modulate interference strength based on ambiguity, and perform quantum measurement for classification.

Specifically:

1. Quantum-inspired Embedding Module: Integrates pre-trained models with quantum encoding to capture rich semantic-syntactic information.
2. Quantum Interference-inspired Dynamic Adaptive Quantum Mechanism: Adjusts interference strength according to sentence-level ambiguity, enabling flexible quantum feature fusion.
3. Quantum Measurement-inspired Sentiment Classification Module: Performs measurements on quantum fused features to predict robust and interpretable results.

Results

Experiments on five benchmarks show that our method achieves SOTA performance with 94.00% average accuracy, outperforming existing quantum-inspired models. Furthermore, additional analyses confirm the method's effectiveness and interpretability.

Roadmap

• Release core model code (QDA)
• Release training and inference scripts
• Share pre-processed datasets (see dataset/)
• Upload pre-trained weights

Datasets

The datasets used in this work are available under dataset/.

citation

If you find this work useful, please verify the status of our paper and cite it:

% BibTeX will be updated upon publication
@article{QDA2024,
  title={QDA: An End-to-End Quantum-inspired Dynamic Adaptive Framework for Textual Sentiment Analysis},
  author={Your Name and Co-authors},
  journal={ArXiv/Journal Name},
  year={2024}
}