Measurement Of Technological Influence:Based On Patent Timeliness And Page Rank Algorithm

Patent, as a unique source of information, can not only reflect the development path of a certain category of technology, but also the distribution of technology within a period of time. Using patent information to research deeper, provides companies with information support in technology innovation strategy, competition analysis, patents value mining, technology research and development. In addition, a company's patent number, patent investment, patent licensing and other information can also reflect its management capacity.Traditional indicators using patent information to evaluate technology influence counts from different sides of patents, may to some extent reflect the importance of companies' technology in the industry. However, current study shows that these indicators, such as patent number, patent citations, the average patent citations, have already shown some drawbacks. After that, evaluation indicators, such as citation index, the reference rate, the relative R & D capabilities, immediate impact indicators, technology cycle time, technology strength, contact science degree, can measure different aspects of enterprise technology innovation processes and results. Unfortunately, these indicators treat equally the reference from authoritative patent and general enterprises. Besides, they can not reflect the extent of reliance from citing enterprises to cited ones. Later initiated ideas to apply the PageRank algorithm to enterprise reference network can make up for these deficiencies. However, it still can not show the different importance between new patents cited and old patents cited, which is remarkable in patent analysis.Thus, in order to effectively identify companies with high technology influence, this study presents an improved evaluation method for technological influence among company reference network on patents. This method innovatively adds the timeliness of patents in the PageRank algorithm application on technological influence measurement. We first present the improved algorithm and show the difference between ours and the traditional one, which is based on PageRank algorithm but without modifying.We then use the patents, details and citations etc. in the field of hydraulic brake from 1976 to 2006 in NBER database, as a sample to do empirical analysis. Our improved method considers five aspects: the reference number of enterprises, citations, their own influence, dependence between enterprises, and patent timeliness. The empirical result shows that, compared with the traditional evaluation method without consideration of timeliness, the improved one tend to give privilege to those whose new patents are cited more, and this can present the importance of new patents. This evaluation method can measure the technological influence of the companies more objectively, and this provides a theoretical basis for technology competitor identification.