2019 Annual Report Intellectual Property Office

V. IPR CREATION AND USE TIPO substantiated its IP Commercialization Education Website by regularly uploading trend analysis of the latest patented technologies, stories about technology transfer, and successful IP commercialization cases. To raise SMEs' IP commercialization awareness and improve their ability in industry analysis and IPR portfolio management, TIPO also made and shared on its website new videos on teaching how to search on the "Global Patent Search System (GPSS)" and "Taiwan Patent Search System (TWPAT) I & II." TIPO also made available the slides from seminars on "Industry analysis based on Patent Search" and "Strategies and practices of global patent search". In addition, a technology matching platform was established for patentees to log in and post information about their patented technologies, which will then be sent to the Taiwan Technology Marketplace (TWTM) for potential IP commercialization. Assistance to the Industry in Commercializing Patents The "IP Corner for SMEs" platform provides an interactive Q&A section and the full-text search function to address the problems that the SMEs may face, and the measures taken by authorities required to their resolution. For the SMEs, it is a fast and convenient tool to find resources or be directed to available services in the process of IP-commercialization. In 2019, TIPO added two links on the platform for the SMEs, one is to the "From IP to IPO program" website of the Ministry of Science and Technology (MOST) and the other is to the "Trade promotion" website of the Bureau of Foreign Trade (BOFT). IP Corner for SMEs h t t ps : / / pcm. t i po . gov. tw/SME/ index.html IP Corner for SMEs 02 Patent Examination Trend Analysis Using keywords in three categories (AI techniques, AI application fields and AI functional applications), this study searched a total of 1,567 AI-related invention patent applications filed in 2018. The connection between these applications and deep learning, employed functionalities, areas of usage, as well as applicants' business sectors were analyzed case by case. The study also looked at examination opinions and results of the 517 AI-related inventions whose OAs had been issued. In AI techniques, neural networks (deep learning) were the top core technologies applied in AI- related inventions (487 inventions, or 31% of all AI-related inventions). However, over half of these AI-related invention applications were not directly related to neural networks. As to fields of application, "Industry and manufacturing" took the lead (279 inventions or 24%) and surpassed the percentage globally. Among AI functional applications, computer vision is the dominant field, which is in line with global trends. Below is the cross-analysis of functional applications and fields of application. The Study on AI-related Patent Applications in Taiwan and Common Reasons for Rejection 62