<p>Sonnet is the traditional form of poetry that presents idea, a message, or an emotion in a semantic way with a rhythmic pattern. Recent advancements in Deep Learning (DL) have made it possible to fuse classical poetry with contemporary technology. This paper introduces a DL-based system for the automatic generation of alluring sonnets. The model is specifically tailored to learn the style and syntax of the Shakespearean sonnets. The neural network, Long Short-Term Memory (LSTM) model, is utilized to learn the structure of a sonnet. To have a thematic corpus for the sequences of sonnets, the cutting-edge Bidirectional Encoder Representations from Transformers (BERT) is exploited. During supervised training, the system identifies lines, sequences, and rhythmic patterns. The model learns patterns and their relationships from a dataset of 52 sonnets. To follow the standard syllables and rhymes of the sonnets, the libraries of PyPhen, Syllables, and Pronounce are utilized. The model’s performance is assessed in terms of poetic quality, coherence, meaningfulness, and grammaticality. Although a little semantic coherence was observed in the generated sonnets, the structure, grammar, and poeticness of the sonnets were up to the mark. The evaluation results indicate that the model can effectively generate stylistically sound sonnets.</p>