Standard-essential Patent Prediction: Discussion of Patent Standardization Time

Abstract

Standard-essential patents are crucial tools in international trade competition, and standard-essential patent prediction is becoming competition point among countries. However, existing standard-essential patent prediction methods cannot adequately use various patent features, and cannot provide verifiable predicted results on whether a valid patent can be a standard-essential patent or not. To address these issues, this paper proposes a dynamic standard-essential patent prediction model. In our model, we partition different time slots for each patent to extract the dynamic features in each time slot of the patent, and the static features of the patent are extracted. Then we dynamically predict whether the patent can be a standard-essential patent in the next few years or not. The experimental results show that our model significantly outperforms the baseline models in the values of accuracy, precision, recall and F1.