Research on a complex network and online review data-driven product innovation design

As a significant conduit for online word-of-mouth, product online reviews (PORs) play a vital function for businesses and prospective customers. However, enormous and diversified comment data has resulted in a severe information overload for users, making it impossible for organizations and consumers to make quick, sensible decisions based on complex data. In this research, PORS content is the node, the semantic similarity between content is the weight of links, and, in conjunction with the idea of a complex network (CN), a PORs network is formed and a framework for the production of data-driven (DD) design concepts is given. The optimization target is determined based on the clustering results, and the feature coding of the optimization target is performed to replicate the psychological assessment mechanism of users. The research demonstrates that when the novelty threshold is set to 0.5, the F-score reaches its maximum value of 0.935%, indicating that the algorithm performs at its peak. When the novelty threshold is set at 0.8, the algorithm's accuracy achieves 0.954%. The practicability and efficacy of the proposed approach are examined.