摘要
[目的/意义]提出一种利用专利信息对特定领域微观层面的技术进行快速预测分析的方法,为企业技术人员研发活动及政府部门的政策制定提供参考。[方法/过程]首先利用S曲线判断该领域是否处于成长期,然后采用多维尺度分析和K均值聚类相结合的方法进行技术聚类并利用IPC号排名对聚类后的技术进行定义,最后通过设计的技术评估指标体系对技术集群在每个指标下的值进行排名,利用孔多塞(Condorcet)投票法选择出有发展前途的技术。[结果/结论]将该预测方法应用于当下亟待指明技术发展方向的中药动物药领域,以2006—2010年为研究区间,预测的结果表明与珍珠或珍珠母类的海洋类中药相关的技术在未来最有发展前途,与国家大力发展海洋经济的战略相契合,并且该预测结果也得到了2011—2015年数据的验证,证明了预测方法的有效性。
[Purpose/significance] This paper presents a method for rapid prediction and analysis of the technology at the microscopic level in a specific field by using patent information,which provides references for RD activities of enterprises＇ technical personnel and for policy making in government departments. [Method/process] Firstly,the S curve is used to determine whether the field is in the growth phase. Next,the technology is clustered by using the combination of multidimensional scaling and K-means clustering and then it is defined by the IPC number ranking. Finally,through the designed technical evaluation indexes system,the paper ranks the values of technical clusters under each index,and uses Condorcet voting to select promising technologies. [Result/conclusion] This prediction method is applied to the field of Chinese medicine and animal medicine in order to find out the technical development direction from 2006 to 2010. The forecast result shows that the technology related to marine Chinese medicine of pearl or nacre has the most promising future,which is in line with China＇s strategy of vigorously developing the marine economy,and also is verified by the data from 2011 to 2015. The results prove the effectiveness of the proposed forecasting method.
出处
《情报理论与实践》
CSSCI
北大核心
2018年第8期79-84,共6页
Information Studies:Theory & Application
基金
辽宁省教育厅高校基本科研项目"基于技术创新的中药企业国际竞争力提升路径研究"的成果之一
项目编号:2017WZD04
关键词
专利信息
专利分析
技术预测
多维尺度分析
孔多塞投票法
patent information
patent analysis
technology forecasting
multidimensional scaling
Condorcet method
