In interpreting user opinion within today’s digital atmosphere, sentiment analysis and emotion detection are crucial. This paper focuses on the application of advanced Natural Language Processing techniques in analyzing and classifying emotions based on customer ratings. The study draws on more than 3000 Amazon Alexa reviews to create a rich source of emotional expressions. Real-world Sentiment and Emotion Patterns. The present study conducts experiments on real-world data, revealing sentiment and emotion patterns while providing guidance for future work and further directions in multimodal emotion detection, as well as fine-tuning pretrained models toward specific fields of application. Precision, recall, accuracy, and F1-score measures are used to evaluate the performance of these three models: LSTM, BiLSTM, and BERT.