SceEmoNet: A Sentiment Analysis Model with Scene Construction Capability
Open Access
- 2 August 2025
- journal article
- research article
- Published by MDPI AG in Applied Sciences
- Vol. 15 (15) , 8588
Abstract
How do humans analyze the sentiments embedded in text? When attempting to analyze a text, humans construct a “scene” in their minds through imagination based on the text, generating a vague image. They then synthesize the text and the mental image to derive the final analysis result. However, current sentiment analysis models lack such imagination; they can only analyze based on existing information in the text, which limits their classification accuracy. To address this issue, we propose the SceEmoNet model. This model endows text classification models with imagination through Stable diffusion, enabling the model to generate corresponding visual scenes from input text, thus introducing a new modality of visual information. We then use the Contrastive Language-Image Pre-training (CLIP) model, a multimodal feature extraction model, to extract aligned features from different modalities, preventing significant feature differences caused by data heterogeneity. Finally, we fuse information from different modalities using late fusion to obtain the final classification result. Experiments on six datasets with different classification tasks show improvements of 9.57%, 3.87%, 3.63%, 3.14%, 0.77%, and 0.28%, respectively. Additionally, we set up experiments to deeply analyze the model’s advantages and limitations, providing a new technical path for follow-up research.Keywords
Funding Information
- Doctoral Research Foundation of Xinjiang Normal University (XJNUZBS2404, XJEDU2025P068)
- Scientific Research Projects of Basic Scientific Research Business Expenses for Universities in Xinjiang Uygur Autonomous Region (XJNUZBS2404, XJEDU2025P068)
- Talent Projects of the Third Batch of the “Tianchi Talent” Introduction Plan in Xinjiang Uygur Autonomous Region (XJNUZBS2404, XJEDU2025P068)
This publication has 42 references indexed in Scilit:
- MCred: multi-modal message credibility for fake news detection using BERT and CNNJournal of Ambient Intelligence and Humanized Computing, 2022
- Compositional coding capsule network with k-means routing for text classificationPattern Recognition Letters, 2022
- Gated attention fusion network for multimodal sentiment classificationKnowledge-Based Systems, 2022
- AMFB: Attention based multimodal Factorized Bilinear Pooling for multimodal Fake News DetectionExpert Systems with Applications, 2021
- Financial causal sentence recognition based on BERT-CNN text classificationThe Journal of Supercomputing, 2021
- Multimodal sentimental analysis for social media applications: A comprehensive reviewWIREs Data Mining and Knowledge Discovery, 2021
- Fusion of heterogeneous attention mechanisms in multi-view convolutional neural network for text classificationInformation Sciences, 2021
- Exploring deep learning approaches for Urdu text classification in product manufacturingEnterprise Information Systems, 2020
- Empirical evaluation of deep learning models for sentiment analysisJournal of Statistics and Management Systems, 2019
- Deep learning for real-time social media text classification for situation awareness – using Hurricanes Sandy, Harvey, and Irma as case studiesInternational Journal of Digital Earth, 2019