Article Preview
TopIntroduction
The linkage and interaction between science and technology (S2T) have been of great interest to researchers (Contopoulos et al., 2003; Malva et al., 2015; Ke, 2019; Ke, 2020). Translational knowledge, which bridges the gap between science and technology, has been shown to play a crucial role in driving innovation (Contopoulos et al., 2003; Contopoulos et al., 2008; Weber et al., 2013). For example, translational science in Biomedicine is a key field that translates basic scientific discoveries (“bench”) into clinical applications (“bedside”) (Ke, 2019). The focus of this study is on the diffusion of knowledge from science to technology within the context of translational research. Technology development absorbs scientific knowledge through various means, including citing scientific articles, direct translation, and indirect translation from other patents. Indirect translation is particularly important for accessing scientific impact and passing on scientific knowledge to future technologies (Jiang & Zhuge, 2019). However, most existing studies focus solely on direct translation between science and technology, such as analyzing direct citation between articles and patents (Ke, 2019; Ke, 2020; Marx, 2020; Marx, 2022), while neglecting the indirect citations that account for most of the S2T translation, especially for basic science knowledge (Narin, 1997). Indirect translation involves many nodes and edges from science to technology, similar to “long ties” in social networks. In the past, long ties were widely assumed to be weak in social emotions and information functions. However, recent studies have proposed that long ties bridging different communities play crucial roles in spreading information and integrating relationships in population-scale social networks (Alstyne, 2011; Lyu et al., 2022; Park et al., 2018). Long ties are more persistent and diverse and nearly as strong as ties embedded within a small circle of nodes. Marra et al. (2015) observed direct and indirect citations on patent life and found that indirect citations account for more complex knowledge flows within the innovation network. Inspired by relevant studies, we hypothesize that the long and indirect S2T translational path has surprising power in innovation. The research question of this study is how direct and indirect translation distributes within the innovation system and how well their knowledge flow effects are.
In this study, we aim to extract S2T translation patterns by analyzing patent-article citation networks. Our primary focus is on understanding the patterns of knowledge diffusion and absorption from science to technology, and we believe that citation is a suitable proxy for this purpose (Ke, 2020). Patents absorb knowledge from scientific articles by citing them. Second, patent-article citations are clearer and more accurate in reflecting knowledge absorption than article-article citations. This is because one article may cite dozens of other articles, and it is difficult to claim that the cited article contributed substantively to the research of the citing article due to the complicated motivations and emotions involved (Teplitskiy et al., 2022). However, when a patent application cites a few articles, it is more likely that the knowledge from these articles acts as the key background technology (Li et al., 2017). Thus, we apply citation generations to the patent-article citation network to exert direct and indirect S2T translation (Hu et al., 2011). To precisely measure the translational path length from science to technology, we further define the translation generation (TG) as a simple and powerful indicator. The TG of a translation event is the number of citation generations needed from an article to a patent and the TG of a group of translations is the mean of the individual translation events. Finally, we measure the translational effects from three aspects: translational lag (TL), translational distance (TD), and translational intensity (TI).