基于因子分析的专利质量评估模型构建

目的:建立一个相对科学、可操作性强的专利质量评估模型,促进专利转化。方法:通过文献回顾和专家咨询,纳入18个较易量化的专利质量指标,再检索“智慧芽”专利网站,以是否发生权力转移或许可为标准,随机选取高质量和低质量专利各250条构成建模数据库,运用因子分析方法,建立评估模型。随后检索并随机选取高质量和低质量专利各100条构成验证集数据库,评价模型对专利质量评估的准确性。结果:相关性检验显示,18个指标之间存在显著相关性,适合因子分析。经对比发现,建立的两个评估模型中,模型二对专利的分选能力略高于模型一。结论:构建的专利质量评估模型筛选能力较强,评估准确性较高,对提高专利质量、促进成果转化有一定的作用。

Establishment of patent quality evaluation model based on factor analysis

Objective To establish the scientific evaluation model of patent quality with a good operability for promo-ting the transformation of patents. Methods Eighteen easily quantified indicators of patent quality were included according to literature review and expert consultation. The database model was established with the right transfer or license as its gold standard, which includes of 250 high quality patients and 250 low quality patents randomly selected from the "PatSnap Website". The evaluation model of patent quality was established by factor analysis. The verification set database was developed, which includes 100 high quality patents and 100 low quality patents randomly selected from the "PatSnap Website" and was used to assess the accuracy of patent quality evaluation model. Results Correlation analysis showed that the 18 indicators of patent quality were closely related, which could thus be used in factor analysis. Comparative analysis displayed that the screening capability of patent quality evaluation model was higher than that of database model. Conclusion The screening capability and accuracy of the evaluation model of patent quality are rather high. The evaluation model of patent quality can improve the patent quality and promote the transformation of patents.