This paper introduces a recurrent neural network (RNN) framework for text classification using multi-task learning. The authors propose three different models that share information across multiple related tasks to improve performance. The first model shares a single LSTM layer across all tasks, while the second model assigns a separate LSTM layer to each task but allows information flow between them. The third model also assigns a separate LSTM layer to each task but introduces a bidirectional LSTM layer to capture shared information. The models are trained jointly on multiple text classification tasks, and the results show that joint learning of related tasks significantly improves performance compared to learning them separately. The proposed models outperform state-of-the-art baselines on four benchmark datasets, demonstrating the effectiveness of multi-task learning in natural language processing tasks.This paper introduces a recurrent neural network (RNN) framework for text classification using multi-task learning. The authors propose three different models that share information across multiple related tasks to improve performance. The first model shares a single LSTM layer across all tasks, while the second model assigns a separate LSTM layer to each task but allows information flow between them. The third model also assigns a separate LSTM layer to each task but introduces a bidirectional LSTM layer to capture shared information. The models are trained jointly on multiple text classification tasks, and the results show that joint learning of related tasks significantly improves performance compared to learning them separately. The proposed models outperform state-of-the-art baselines on four benchmark datasets, demonstrating the effectiveness of multi-task learning in natural language processing tasks.