关键词: 深度学习, 专利价值评估, 半监督学习, 大语言模型, 集成学习

Abstract: Patent value evaluation is a crucial tool for optimizing resource allocation and guiding intellectual property strategy decisions. Traditional manual evaluation methods are limited by subjective expert experience and low evaluation efficiency, making it difficult to meet the large-scale demands of enterprises in the digital economy era. In recent years, machine learning technologies, with their powerful high-dimensional feature extraction capabilities, have provided a feasible technological pathway for innovating patent value evaluation paradigms. However, existing studies mainly focus on small-scale models within a single technical dimension and have yet to fully explore the potential of Large Language Models (LLMs) in quantifying value indicators. Moreover, current methods struggle to handle cases where some patent data indicators are missing. To addresses the challenges of processing unstructured textual data and incomplete patent value indicators in the patent database of power grid enterprises, this paper proposed a deep learning-based patent value evaluation method for power grid enterprises. The approach primarily involved: utilizing Large Language Model (LLM) technology to process unstructured textual information in power grid enterprise patents; adopting a semi-supervised learning paradigm to expand the labeled patent database used for training; and employing ensemble learning techniques to train the model on the power grid enterprise patent database and conduct patent value evaluation. Empirical results demonstrate that the proposed method can efficiently evaluate patent value with minimal evaluation error. This research provides a new methodological framework for intelligent patent analysis and has practical significance in supporting intellectual property strategy decisions.

Key words: Deep Learning, Patent Value Evaluation, Semi-Supervised Learning, Large Language Models, Ensemble Learning