Hybrid Model Based on Recurrent Neural Network and Gated Recurrent Unit for Text Processing
Download| Volume 7 Issue 1, 2026 | |
|---|---|
| Author(s): |
Areej Fatima Department of Information Technology, The University of Haripur, Pakistan, areejfatima62880@gmail.com Muhammad Usama Yousaf* Faculty of Information and Communication Technology, Universiti Tunku Abdul Rahman, Malaysia, its.usamayousaf@gmail.com Ayesha Saddiqua Institute of Computing, Kohat University of Science and Technology, Pakistan, kim.ayeshasaddiqua@gmail.com Mahnoor Izzat Faculty of Information and Communication Technology, Universiti Tunku Abdul Rahman, Malaysia, mahnoorizat187@gmail.com Aiman Sohaib Institute of Computing, Kohat University of Science and Technology, Pakistan, aimansohaib223@gmail.com Sana Hassan Faculty of Business and Finance, Universiti Tunku Abdul Rahman, Malaysia, sanahassan@1utar.my Muhammad Adnan Kaim Khani Faculty of Information and Communication Technology, Universiti Tunku Abdul Rahman, Malaysia, adnankk12@1utar.my |
| Abstract | This work proposes a hybrid deep learning model combining Recurrent Neural Networks (RNN) and Gated Recurrent Units (GRU) for improved text processing. While RNNs capture sequential dependencies and GRUs address vanishing gradient issues through gating mechanisms, each has limitations when used independently. The hybrid model leverages the strengths of both architectures to enhance memory retention, contextual understanding, and overall performance on text-related tasks. The proposed model is evaluated on text classification tasks and compared against standalone RNN, GRU, and LSTM models. Experimental results show that the hybrid approach achieves higher accuracy and more coherent, context-aware outputs while reducing training time. By combining RNN’s sequence modeling capability with GRU’s efficient long-range dependency handling, the model overcomes common challenges in memorizing long sequences. The improved performance demonstrates practical benefits for applications such as chatbots, automated summarization, and sentiment analysis, where accurate and contextually relevant text processing is critical. The hybrid model also shows better efficiency in managing sequential data across multiple domains compared to traditional sequential models. Although transformer-based models currently lead in state-of-the-art benchmarks, this hybrid approach offers a computationally lighter alternative with competitive results, making it suitable for scenarios with limited resources. The dataset was split into multiple training and testing ratios to validate robustness, and evaluation across several performance metrics confirms that the proposed model outperforms individual RNN, GRU, and LSTM baselines. The results indicate that combining complementary recurrent architectures can yield a more effective and efficient solution for natural language understanding tasks. |
| Keywords | Recurrent neural network, Gated recurrent unit, Long Short Term Memory, Text classification, Sentiment analysis. |
| Year | 2026 |
| Volume | 7 |
| Issue | 1 |
| Type | Research paper, manuscript, article |
| Recognized by | Higher Education Commission of Pakistan, HEC | Category | Journal Name | ILMA Journal of Technology & Software Management | Publisher Name | ILMA University | Jel Classification | -- | DOI | - | ISSN no (E, Electronic) | 2790-590X | ISSN no (P, Print) | 2709-2240 | Country | Pakistan | City | Karachi | Institution Type | University | Journal Type | Open Access | Manuscript Processing | Blind Peer Reviewed | Format | Paper Link | https://ijtsm.ilmauniversity.edu.pk/arc/Vol7/i1/pdf4.pdf | Page | 35-41 |