"50,000 Patent Information in 4 Hours"... LG Innotek Builds AI Analysis System

[Asia Economy Reporter Kim Heung-soon] LG Innotek announced on the 6th that it has established a system that analyzes tens of thousands of patent information big data in a short time using artificial intelligence (AI).

According to LG Innotek, this system, called the 'AI Patent Information System,' processes tasks about 900 times faster than humans. For example, analyzing 50,000 patent information cases would require five people working 8 hours a day for a total of 100 days, but this system can complete the task in just over 4 hours. The company explained that this enables a shift from the traditional method of creating patents after development to establishing research and development (R&D) strategies through patent analysis.

The patent information system is used for protecting intellectual property and managing risks related to research outcomes, such as checking whether there are similar patents for new products or new technologies, or whether other patents are infringed. LG Innotek developed an AI sentence recognition model with LG AI Research Institute to quickly process patent information. This model automatically analyzes the text and context of patent documents in real time to rapidly extract key information such as patent application fields, problems to be solved, and solutions. Additionally, the AI continuously learns classification criteria for technical data and can automatically classify tens of thousands of patent information cases into detailed technical units.

LG Innotek plans to utilize the AI Patent Information System for product and technology development as well as mid- to long-term R&D strategy planning. Kang Min-seok, Vice President and Chief Technology Officer (CTO) of LG Innotek, said, "Patent information is a leading indicator that shows the flow of future industries and markets. In the future, companies that effectively utilize patent big data accumulated worldwide will hold technological leadership," adding, "We will present technologies and products that impress customers through strategic R&D using AI, big data, and digital transformation (DX)."