关键词: 专利相似度, IPC分类, 突破性技术发明, 识别方案, 纳米技术
Abstract: With the rapid increase in the number of patent applications in China, how to measure and identify breakthrough inventions with high technology and economic value has become a hot issue. Since 2011, the number of invention patent applications in China has ranked first in the world for eight consecutive years. However, how to identify breakthrough inventions through ex ante and ex post indicators remains a problem in academia. Thus, it is of significance to identify the invention patents with major breakthroughs and turn them into productivity in China.
Considering the general lack of citation information in Chinese patents, this paper draws on the definitions and identification criteria of breakthrough inventions by Dahlin and Behrens (2005) to construct the patent-to-patent similarity index with the IPC information of patents. Considering both the ex ante and ex post indicators of patents, we introduces the time dimension to measure the novelty, uniqueness, and impact of patents by comparing the similarity of patents in the past, current and future periods, so proposing a comprehensive identification solution for breakthrough inventions. In order to test the validity of our identification solution, we choose the field of nanotechnology as an example. As one of the fastest growing and widely influential emerging science and technology fields in the world, nanotechnology is a typical interdisciplinary technology field. The cross-domain attributes of nanotechnology determine the wide range of IPC classifications involved, which is suitable for our research using patent IPC classification and is more useful to reveal the invention characteristics of technologies in the emerging interdisciplinary domains. Then, we conduct an empirical test by using the granted invention patents in nanotechnology from the United States Patent and Trademark Office (USPTO) in 1989-2010.
Our conclusions are as follows. First, we find that it is feasible and operational to use IPC classification similarity to identify breakthrough inventions. Specifically, we can identify 6.23% and 5.06% of nanotechnology patents as breakthrough inventions with the similarity indicators based on IPC four- and six-digit classifications. Second, by comparing with the breakthrough inventions identified based on patent citation, this paper verifies the validation and difference of the breakthrough inventions identified based on patent IPC classification similarity. This paper finds that there is a significant positive correlation between breakthrough inventions identified based on patent similarity and based on patent forward citations. At the same time, however, this paper finds that only 0.5% of the results obtained by the two types of identification methods are the same, indicating that it may be biased by purely relying on patented citation data to identify breakthrough inventions in the past, so the identification solution based on IPC classification similarity is a useful supplement. Finally, through the comparative study of the source characteristics at the invention-, inventor- and organizational-level, this paper further validates the effectiveness and difference of the breakthrough inventions identified based on patent IPC classification similarity. The breakthrough inventions based on patent similarity and patent citation shows basically consistent results in inventor and organizational characteristics, indicating the validation of the new identification solution based on patent IPC similarity in the inventor- and organizational-level. At the same time, however, this paper also finds that the two types of identification methods show inconsistent results in the characteristics of the invention′s knowledge source, indicating that the previous research results on the source characteristics of the invention may be affected by the breakthrough invention identification method. Thus, we suggest that future research on the source of breakthrough inventions needs to pay attention to different identification methods.
Key words: patent similarity, IPC classification, breakthrough invention, identification method, nanotechnology