The Matic Analysis of Innovative Drug Policies Based on LDA Modeling

Objective Based on LDA topic modeling, 423 policy texts in the field of innovative drugs in China are subject to content topic mining and analyze the features and focus of the policies. Methods Based on the reading of 423 innovative drug policies and unsupervised machine learning of the policy texts using LDA topic model to find the potential themes of the texts and categorize them. Results The study found that the innovative drug policy texts show obvious life cycle characteristics, and the current stage is in the policy deepening period. Through the analysis, the policy content themes can be classified into three major categories, namely, drug patents and R&D incentives for innovative drugs, health insurance policies and market access for innovative drugs, and policy support for innovative drugs and industrial development. Among them, drug patents and access to innovative drugs in health insurance are the most prominent hot topics. With the passage of time, each theme shows different evolution trends.