Transformative Conversational AI: Sentiment Recognition in Chatbots via Transformers

Recent years have seen a dramatic increase in the use of conversational artificial intelligence (CAI) for both academic and commercial applications, primarily in the context of chatbots and AI virtual assistants. The user’s engagement produces a human like responses. However, the capacity to discern sentiments and respond adequately is one of the major difficulties faced by conversation systems. In the present study, we propose a transformer-based framework for a sentiment-aware chatbot. The suggested transformer is a neural network architecture that is highly parallelizable and solely dependent on the self-attention mechanism. A transformer model controls variable-sized input using stacks of self-attention layers rather than deep neural networks or CNNs. In this manner, language creation is carried out using the cutting-edge pre-trained model CTRL, which can easily adapt to various pre-trained models without needing architectural adaptations. Our model was trained using the DailyDialogues dataset and was evaluated using automated metrics. Findings from experiments confirm that, in terms of content quality and emotion perception, our suggested technique works better than cutting-edge baselines.