Research on the Graded Classification Management System Based on Multi-Dimensional Patent Value Evaluation

Based on patent data from the State Grid Information & Telecommunication Group, this study proposes the SGScore model. The model evaluates patents from multiple dimensions, including technical fields, legal status, historical transformation, patent citations, and patentee attributes, integrating a patent scoring rule to construct a comprehensive "4+1+N" evaluation framework. Through data cleaning and statistical analysis, the study utilizes the model to quantitatively assess patent value and explores graded classification management of patents. Results show that the SGScore model accurately identifies high-value patents, optimizes corporate patent resource allocation, and provides scientific support for technological planning and market expansion. This research offers a theoretical foundation and practical tools for refined management of corporate patent assets, advancing the development of patent value evaluation methods.