Research of Density Based Spatial Clustering of Applications with Noise on TCM Patent Law Evaluation

Abstract

DBSCAN (density-based spatial clustering of applications with noise) algorithm is a spatial clustering algorithm based on density. It has a good effect in dealing with discrete data sets. As an important branch of data mining, this clustering analysis method studies the classification of data objects in broad application prospects such as pattern recognition, image processing, market research, and life science. Based on the DBSCAN algorithm, this paper designs a simple formula to solve the value of parameters EPS and MinPts collects the relevant legal evaluation index data of traditional Chinese medicine patent, and finally establishes an evaluation model of traditional Chinese medicine patent law based on DBSCAN. The average accuracy of the model reaches 91.97%, which is superior to other domestic evaluation models of traditional Chinese medicine patent law. It is proved that the model is not only reliable in the legal evaluation of traditional Chinese medicine patents, but also superior to other domestic models.