A Robust and Efficient Attention-CNN-BiGRU Framework for Sentiment Analysis in Noisy Social Media Environments
CyberSystem Journal · 24.06.2026 · 3 (1) · 36-46
DOI: 10.57238/csj.2026.1024 ↗Kurz gefasst
Fachartikel in der Zeitschrift CyberSystem Journal, online erschienen am 24.06.2026. Verfasst von Rahman Abid. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The proliferation of social media platforms, particularly X, has revolutionized public communication but poses challenges for sentiment analysis due to informal language characterized by abbreviations, slang, emojis, and creative spellings. This study presents an Attention-Enhanced Convolutional Neural Network–Bidirectional Gated Recurrent Unit (CNN-BiGRU) model, a hybrid architecture designed for both accuracy and computational efficiency. It integrates CNNs for local feature extraction, BiGRUs for capturing sequential dependencies, and an attention mechanism to prioritize sentiment-critical components. Evaluated on Sentiment140 (1.6 million tweets, binary classification) and SemEval-2017 Task 4 (65,000 tweets, three-class classification), the model demonstrates statistically significant accuracy improvements (up to 1.7 percentage points) compared to baseline models, including BERT Base, while requiring substantially fewer parameters and offering potential for faster inference. The model achieves 86.3% accuracy on Sentiment140 and 77.2% on SemEval-2017 Task 4, effectively handling linguistic variability. Its comparatively lightweight design makes it suitable for real-time sentiment analysis scenarios where computational resources may be constrained. This framework offers a robust and scalable solution for analyzing sentiment in evolving online communication, with applications in market research, public opinion monitoring, and crisis management.
Zitieren
APA
Abid, R. (2026). A Robust and Efficient Attention-CNN-BiGRU Framework for Sentiment Analysis in Noisy Social Media Environments. CyberSystem Journal, 3(1), 36-46. https://doi.org/10.57238/csj.2026.1024
Vancouver
Abid R. A Robust and Efficient Attention-CNN-BiGRU Framework for Sentiment Analysis in Noisy Social Media Environments. CyberSystem Journal. 2026;3(1):36-46. doi:10.57238/csj.2026.1024
BibTeX
@article{Abid2026,
author = {Abid, Rahman},
title = {A Robust and Efficient Attention-CNN-BiGRU Framework for Sentiment Analysis in Noisy Social Media Environments},
journal = {CyberSystem Journal},
year = {2026},
volume = {3},
number = {1},
pages = {36--46},
doi = {10.57238/csj.2026.1024}
}
EUVIMED