The application of patent citation analysis to the study of knowledge diffusion has not yet reached firm conclusions. Controversy still flourish on whether the interpretation of citations as ‘paper trails’ left by interpersonal knowledge flow is legitimate. In this paper we have stressed that much of the debate depends on the popularity and the peculiarities of the US patent system, whose limitations may not affect the European one. In particular, many efforts placed on distinguishing between inventors’ and examiners’ citations are pointless when using EPO data. We have also stressed how the same efforts can indeed be judged excessive for the purpose of using USPO data, as long as one recognizes that knowledge of the technical contents of a patent may travel independently from information on the existence of that patent, or from exact references to the relevant documents. Social network analysis can be more decisive, for at least three reasons. First and foremost, because it recognizes that information may travel from inventor to inventor not only directly, but also indirectly, via complex social chains. Second, because inventors, at least in R&D and patent intensive fields, may well represent a ‘community of experts’, that is a meaningful unit of analysis. Third, because a methodology has emerged from the recent literature, which allows us to test the influence of social distance on citation probabilities. When applied to EPO data, that methodology confirms that short social chains of inventors are indeed more likely to generate citations than unconnected inventors.
Knowledge networks from patent data: methodological issues and research targets
BRESCHI, STEFANO;
2004
Abstract
The application of patent citation analysis to the study of knowledge diffusion has not yet reached firm conclusions. Controversy still flourish on whether the interpretation of citations as ‘paper trails’ left by interpersonal knowledge flow is legitimate. In this paper we have stressed that much of the debate depends on the popularity and the peculiarities of the US patent system, whose limitations may not affect the European one. In particular, many efforts placed on distinguishing between inventors’ and examiners’ citations are pointless when using EPO data. We have also stressed how the same efforts can indeed be judged excessive for the purpose of using USPO data, as long as one recognizes that knowledge of the technical contents of a patent may travel independently from information on the existence of that patent, or from exact references to the relevant documents. Social network analysis can be more decisive, for at least three reasons. First and foremost, because it recognizes that information may travel from inventor to inventor not only directly, but also indirectly, via complex social chains. Second, because inventors, at least in R&D and patent intensive fields, may well represent a ‘community of experts’, that is a meaningful unit of analysis. Third, because a methodology has emerged from the recent literature, which allows us to test the influence of social distance on citation probabilities. When applied to EPO data, that methodology confirms that short social chains of inventors are indeed more likely to generate citations than unconnected inventors.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
