论文部分内容阅读
应用马尔可夫链对专利引证网络的行走过程进行建模,运用数理统计的方法衡量知识在技术发展过程中的重要性和价值,进而分析知识继承和发展的情况。论证了知识流的聚合原理。专利引证网络的时间无关性、不可约性以及引证路径单向性满足马尔可夫链收敛的条件,使得专利访问概率转化为马尔可夫链转移矩阵稳态分布。从美国专利商标局的专利数据库中采集了3887项1976—2006年授权的燃料电池专利,从美国国家经济研究局的专利数据集中查询得到22074条专利引证关系,利用大型复杂网络分析工具Pajek构建专利引证网络。实验结果表明:极少数专利构成的知识进化轨迹积聚了整个专利引证网络近1/4的访问概率,符合知识流的表现形式;在知识进化轨迹中知识继承和发展的情况并存,符合知识流的内在特征。
The Markov chain is used to model the walking process of patent citation network. The method of mathematical statistics is used to measure the importance and value of knowledge in the process of technological development. Then the situation of knowledge inheritance and development is analyzed. The polymerization principle of knowledge flow is demonstrated. The time independence, irreducibility and unidirectionality of citation networks satisfy the condition of Markov chain convergence, which translates the probability of patent access into the steady state distribution of Markov chain transfer matrix. 3,887 fuel cell patents licensed from 1976 to 2006 were collected from the USPTO patent database, 22,074 patent citations were obtained from the patent data sets of the US National Bureau of Economic Research, and the patent was built using Pajek, a large and complex network analysis tool Citation Network. The experimental results show that: the trace of knowledge evolution formed by very few patents accumulates the access probabilities of nearly a quarter of the total number of patent citation networks, which accords with the representation form of knowledge flow; the inheritance and development of knowledge coexist in the trajectory of knowledge coexistence with knowledge flow Intrinsic characteristics.