Intelligent Connected Vehicle Industry 2015-2023 Patent and Paper Dataset

This study first screens out core patents and treats them as structured data, using the random forest algorithm to identify their important features; at the same time, unstructured data such as journal articles and project information are represented through a word vector model. Then, through data fusion methods, structured and unstructured data are integrated at the feature level, and the LDA model is applied to identify key core technology topics in the industry, while the patent network method is used to determine cutting-edge patent clusters, comprehensively obtaining cutting-edge key core technology topics.

Finally, based on the identified key core technology results, a study on country distribution is conducted to clarify the international competition situation of key core technologies in the industry.