Structure Analysis On Patent Technological Topics With Knowledge Network:in The Field Of3G Mobile Communication

With the development of system of science and technology, the inventions have changed the national innovation competitiveness in the macro level, influenced the industrial progress in the meso perspective, and affected the firm’s prosperity in the micro scope. So the filing and granted patent are adopted to assess the technology competitiveness of the country and company. At the same time, patent document, the technological knowledge carrier, contains the innovation fruits of researchers. So, how to represent the structure and evolvement of patent technological topics, has to be an important issue that must be solved in the front of researchers.At first, the theoretical frame is built from the concepts of knowledge network and knowometrics, and then the meaning of patent technological topic is defined. After that, we explain that the formulation of patent technological topic is the result of division and recombination of knowledge units, at the same time, the mechanism from technological term to technological topic is analyzed to further study the main path to technological topic. Furthermore, the analysis models, analysis indicators and analysis methods are constructed based on the matrix of co-occurrence technological terms and the matrix of association between technological topics.Secondly, taking the patents of3G mobile communication technology as example, the paper extracts the terms from title and abstract to represent the technological topics. Furthermore, cosine index is implemented to compute the similarity of the technological topics to form co-topics network, in which the node shows the topic and the link means the co-occurrence of topics. Then the paper employs strategic coordinate to reveal the structure and transformation of technological topics, based on the idea of co-occurrence words introduced by Collon. In the co-topic network, the core topics play the important role in the technology development, and they have the tendency of crossover and mergence. With the analysis on the dynamic network, it shows that the changement of the network structure has great relationship with the correalation between the technological topics in different periods. With all above, the evolution of the research content and its direction could be revealed.At last, with the same patent data, we construct the matrix of correlation and the matrix of co-occurrence maximum outer connection for the patent technological topics. Then, the correlation structure and the strong link structure of the technological topics are made analysis from the perspective of relationship. We find that the topics with higher betweenness are most important, with the trend of crossover and convergence, while the ones with higher maximum ounter degree influence the strong link structure. With the time goes on, the betweenness and maximum outer degree of the technological topics are changed, which displays the inherent and update of the topics, introduce of the new topics and new links, the increment of the topics link, and so on.In short, structure analysis on the patent technological topics with knowledge network, can’t fully appear the mechanism of the evolvement, but it could be used to represent the process of the division and recombination of the knowledge units objectively. Also to some extent, the periodicity, correlation and foreseeability of the development of the technological topics are in a clear view. All this could assist the policy-maker in management policy making, project planning and project supporting, in order to promote the patent technology research and technology development.