In the dynamic landscape of e-commerce applications, recommendation technology has established a unique role by aiding users in selecting items or products from a vast array of choices. Nevertheless, when a new user becomes part of the recommender system, it faces a challenge in providing relevant item suggestions due to the absence of prior information about the user and their historical interactions with items. This predicament, known as the ‘cold-start’ problem, remains unsolved, without a definitive solution in place. In recent times, deep learning methods have gained widespread popularity and are increasingly utilized for addressing cold start challenges. Deep learning algorithms need to be used to investigate the results for this problem across multiple fields.