With the advent of Industry 4.0, the competitive environment has changed. Therefore, this research, within the framework of Industry 4.0, aims to identify consumer needs and provide optimization solutions, seeking to develop a new product based on consumer opinions. In this study, the Dirichlet Latent Allocation method was used, which is a statistical approach based on topic modeling and enables the automatic extraction of hidden patterns and topics from large text data. In this study, 6394 user opinions were collected from the Amazon site using the Python programming language and web data mining method. Next, based on these comments, a word cloud of the 50 most important words was drawn. After that, by implementing the Dirichlet latent assignment model in the Python environment, the comments related to each of the product models were analyzed and evaluated using this algorithm. The results showed that 20 topics, including 10 topics related to product features and 10 topics related to services, were of great importance from the consumers’ perspective, and the technical features of new product development were determined. The findings of this research can be used as a practical tool for designers and product development managers in making decisions related to improving design, enhancing quality to the real needs of the market. Finally, using the best-worst method, 10 topics related to product features were ranked; So that the processor, RAM, and screen ranked first to third, and the ports, keyboard, and audio system ranked lowest among product features.