A Trend Analysis Method for IoT Technologies Using Patent Dataset with Goal and Approach Concepts

Abstract

With the recent tremendous advances in Internet of Things (IoT), analyzing technology trends and providing high quality services also play an important role in analyzing big data. Patent data has been on the rise, and the importance of patents also has been increasing to demand a company’s competitiveness and to claim technology rights since patents contain important feature technologies that can be used to analyzing new technology trends and to find competitor companies or technologies. Especially IoT techniques are evolving rapidly and a number of new technologies have emerged. Hence a variety of analysis techniques on inventions patented by specific companies are required. In this paper, a trend analysis method is proposed to extract patterns from patent dataset using ETI relations. The experiment results based on the real patent dataset of IoT domain invented a specific organization show that the characteristics of the organization that have the expansion relations is 98.6 % and the transition relations is 1.4 % among ETI relations. Furthermore, we verify proposed dictionaries are useful for enhancing the accuracy to extract ETI relations from a patent network automatically.