基于机器学习的专利质量评估研究
|
摘要
专利中含有丰富的计量信息,这些计量信息与专利质量之间存在着一定的联系。论文利用3D打印相关的六千多个专利族的十一组专利计量信息,测试了逻辑回归、支撑向量机和神经网络三种机器学习模型在专利质量评估上的表现。神经网络模型获得了较好的实验效果,运用逻辑回归模型可以对准确率或召回率进行单边控制,从而满足特定的应用需要。
Patent contains rich metric information which is related to patent quality. In this study,the information of patent is used to estimate its quality based on three machine learning models. The models include logistic regression,support vector machine and neural network. More than 6,000 patent families in 3 D printing field have been evaluated by these three methods. This study shows that neural network achieves the best result,and the logistic regression can be used in precision and recall control to meet the needs of specific applications.
Patent contains rich metric information which is related to patent quality. In this study,the information of patent is used to estimate its quality based on three machine learning models. The models include logistic regression,support vector machine and neural network. More than 6,000 patent families in 3 D printing field have been evaluated by these three methods. This study shows that neural network achieves the best result,and the logistic regression can be used in precision and recall control to meet the needs of specific applications.
引文
[1]李振亚,孟凡生,曹霞.基于四要素的专利价值评估方法研究[J].情报杂志,2010,29(8):87-90.LI Zhenya,MENG Fansheng,CAO Xia. The evaluation method research of patent value based on the four factors[J]. Journal of Intelligence,2010,29(8):87-90.
[2]Narinf. Technology indicators and corporate strategy[J].Review of Business,1993,14(3):19-23.
[3]Lerner J. The importance of patent scope:an empirical analysis[J]. The RAND Journal of Economics,1994,25(2):319-333:89-95.
[4]Lanjouw J O,Schankerman M. The quality of ideas:measuring innovation with multiple indicators[R]. National Bureau of Economic Research,1999:18-18.
[5]胡启超. BP神经网络在专利价值评估中的应用研究[D].哈尔滨:哈尔滨工业大学,2013:45-58.HU Qichao. Research on application of BP neural network in valuation of patent[D]. Harbin:Harbin Institute of Technology,2013:45-58.
[6]Trappey A J,Trappey C V,Wu C-Y,et al. A patent quality analysis for innovative technology and product development[J]. Advanced Engineering Informatics,2012,26(1):26-34.
[7]Ercan S,Kayakutlu G. Patent value analysis using support vector machines[J]. Soft Computing,2014,18(2):313-328.
[8]ABU-MOSTAFA Y S,MAGDON-ISMAIL M,LIN H-T.Learning from data[M]. AMLBook,2012:4-5.
[9]马廷灿,李桂菊,姜山,等.专利质量评价指标及其在专利计量中的应用[J].图书情报工作,2012,56(24):89-95.MA Tingcan,LI Guiju,JIANG Shan,et al. Patent quality evaluation indicators and their applications in patentometrics[J]. Library and Information Service,2012,56(24):89-95.
[10]HOSMER D,LEMESHOW S. Applied logistic regression[M]. New York:John Wiley&Sons,2004:1-6.
[11]BOSER B E,GUYON I M,VAPNIK V N. A training algorithm for optimal margin classifiers[C]//Proceedings of the Fifth Annual Workshop on Computational Learning Theory,1992:144-152.
[12]李航.统计学习方法[M].北京:清华大学出版社,2012:115-124.LI Hang. Statistical learning method[M]. Beijing:Tsinghua University press,2012:115-124.
[13]邓乃扬,田英杰.数据挖掘中的新方法—支持向量机[M].北京:科学出版社,2004:164-223.DENG Naiyang. TIAN Yingjie. New method in data ming support vector machine[M]. Beijing:Science press,2004:164-223.
[14]HAGAN M,DEMUTH H,BEALE M. Neural network design 2nd[M]. Boston:Pws Pub,2014:14-36.
[15]宗成庆.统计自然语言处理(第二版)[M].北京:清华大学出版社,2013:202-207.ZONG Chengqing. Statistical natural language processing[M]. The second edition. Beijing:Tsinghua University press,2013:202-207.
[2]Narinf. Technology indicators and corporate strategy[J].Review of Business,1993,14(3):19-23.
[3]Lerner J. The importance of patent scope:an empirical analysis[J]. The RAND Journal of Economics,1994,25(2):319-333:89-95.
[4]Lanjouw J O,Schankerman M. The quality of ideas:measuring innovation with multiple indicators[R]. National Bureau of Economic Research,1999:18-18.
[5]胡启超. BP神经网络在专利价值评估中的应用研究[D].哈尔滨:哈尔滨工业大学,2013:45-58.HU Qichao. Research on application of BP neural network in valuation of patent[D]. Harbin:Harbin Institute of Technology,2013:45-58.
[6]Trappey A J,Trappey C V,Wu C-Y,et al. A patent quality analysis for innovative technology and product development[J]. Advanced Engineering Informatics,2012,26(1):26-34.
[7]Ercan S,Kayakutlu G. Patent value analysis using support vector machines[J]. Soft Computing,2014,18(2):313-328.
[8]ABU-MOSTAFA Y S,MAGDON-ISMAIL M,LIN H-T.Learning from data[M]. AMLBook,2012:4-5.
[9]马廷灿,李桂菊,姜山,等.专利质量评价指标及其在专利计量中的应用[J].图书情报工作,2012,56(24):89-95.MA Tingcan,LI Guiju,JIANG Shan,et al. Patent quality evaluation indicators and their applications in patentometrics[J]. Library and Information Service,2012,56(24):89-95.
[10]HOSMER D,LEMESHOW S. Applied logistic regression[M]. New York:John Wiley&Sons,2004:1-6.
[11]BOSER B E,GUYON I M,VAPNIK V N. A training algorithm for optimal margin classifiers[C]//Proceedings of the Fifth Annual Workshop on Computational Learning Theory,1992:144-152.
[12]李航.统计学习方法[M].北京:清华大学出版社,2012:115-124.LI Hang. Statistical learning method[M]. Beijing:Tsinghua University press,2012:115-124.
[13]邓乃扬,田英杰.数据挖掘中的新方法—支持向量机[M].北京:科学出版社,2004:164-223.DENG Naiyang. TIAN Yingjie. New method in data ming support vector machine[M]. Beijing:Science press,2004:164-223.
[14]HAGAN M,DEMUTH H,BEALE M. Neural network design 2nd[M]. Boston:Pws Pub,2014:14-36.
[15]宗成庆.统计自然语言处理(第二版)[M].北京:清华大学出版社,2013:202-207.ZONG Chengqing. Statistical natural language processing[M]. The second edition. Beijing:Tsinghua University press,2013:202-207.