A Patent Data Meta-Path-based Technological Risk Prediction Method

Although many countries have achieved rapid advancements in science and technology, they still rely on other nations in certain key technological fields. Effective prediction of technological risks is vital for national economic growth. Recently, countries have placed a strong emphasis on technological security. Technological risks arise from factors such as technology itself, its environment and management. Predicting these risks helps nations, enterprises and institutions achieve independent and controllable technology. It also enables effective management and control of potential risks. Risk prediction research encounters challenges such as technological diversity and the complexity of technological risks. To address the above issues, we propose a patent data meta-path-based technological risk prediction method. This method takes into account both technological diversity and the complexity of technological risks when predicting risks. Technological risk prediction involves analyzing potential risks and their degrees of severity in technological development. Compared to the baseline methods using our collected patent data, our method performs better in evaluation metrics, demonstrating its applicability in predicting technological risks.