Technology Topic Extraction and Analysis Based on BERTopic ChatGPT4o

Technology theme identification is an important component of technology management, which has significant implications for the development strategies of both enterprises and governments. Patent texts contain rich technical details, making them one of the best data resources for technology recognition. However, existing technology topic recognition methods cannot completely eliminate human intervention and still rely on experts to manually evaluate the modeling results of technology topics and interpret and summarize their connotations. Therefore, to address this issue, this paper proposes a technology topic extraction model based on BERTopic ChatGPT4o fusion, which extracts technology topics from patent texts and uses a large model approach to automatically evaluate and interpret the identified technology topic content to generate topic labels. On this basis, taking the invention patent in the field of artificial intelligence as an example, the effectiveness of the method is verified. Experimental verification shows that the method proposed in this paper has significantly improved the accuracy of technical topic extraction and reduced manual intervention.