Using the data mining method to assess the innovation gap: A case of industrial robotics in a catching-up country

It is critical for "catching-up" countries to narrow innovation gaps with developed countries by developing emerging industries. This research introduces a data-mining based method to systematically assess the national innovation gap that is specifically for emerging industries. The method examines the five key attributes of emerging industries, including the ownership of platform technologies, globalization intention, international knowledge position, university-industry linkage, and cross-disciplinary technology development. In particular, this method combines data-mining with experts' knowledge to build patent-training examples, and then uses a support vector machine-based classifier to single out all high-quality patents for each innovation attribute. Based on the selected high-quality patents, the authors utilize a factorial design analysis to systematically evaluate the innovation gap between countries. This method can significantly reduce measurement bias of traditional single patent indicators. In addition, it also can robustly adjust measuring weights in response to the specifics of each innovation attribute, while traditional multi-attribute evaluation methods cannot. As a result, this research empirically shows that China' industrial robot sector has apparent innovation gaps compared to developed economies, specifically in university-industry linkage, cross-disciplinary competence, and globalization intention, and this calls for the attention of policy makers and industrial experts. (C) 2017 The Authors. Published by Elsevier Inc.