摘要
[目的/意义]识别新颖专利代替由一组关键词代表的专利空白,改善技术机会识别过于主观的缺陷。[方法/过程]利用一种基于系统流程的定量方法识别专利的新颖程度。通过基于密度的局部离群点(DLOF)算法识别出新颖专利,利用技术范围指标与同类专利数量指标构建研发组合标识图。[结果/结论 ]研究结果表明,基于专利的新颖性研发组合标识图可以准确地识别出新颖专利,为技术研发提供借鉴。
[Purpose/significance] The meanings of potential technology opportunities become more explicit by identifying anomaly patents rather than patent vacancies that are usually represented as a simple set of keywords. [Methods/Process] We propose an approach to detecting anomaly patents based on systematic processes and quantitative outcomes. Density-based Local Outlier(DLOF) algorithm is used to identify novelty patents, then use scope of technology index with amount of similar patents index structure Anomalyportfolio patent map. [Results/Conclusion] Research results show that novelty-focused patent mapping for technology opportunity analysis can accurately identify the novelty patents and provide a reference for the technology research and development.
出处
《知识管理论坛》
2016年第4期276-282,共7页
Knowledge Management Forum
基金
吉林省教育厅"十二五"社会科学研究规划课题"基于专利分析的吉林省生物制药产业创新发展战略与实施路径研究"(项目编号:2014B019)的研究成果
关键词
技术机会分析
新颖专利
专利新颖性研发组合标识图
文本挖掘
局部异常因子
technology opportunity analysis
novelty-focused
novelty-focused patent identification map
text mining
local outlier factor