SDK and IBM Jointly Develops AI-Based Patent Interpretation Support System
―High Readability Helps Engineers to Reduce Time for Screening of Patent Information―
Showa Denko K.K.
April 10, 2019
Showa Denko (SDK) (TOKYO:4004) and IBM Japan, Ltd. has jointly developed a “Patent Interpretation Support System,” which helps engineers to conduct effective and efficient screening of patent information. SDK will start to utilize this system in the whole company from this July, aiming to help engineers to reduce time spent on comprehension and interpretation of patent information.
The “Patent Interpretation Support System,” jointly developed by SDK and IBM Japan this time, adopted “IBM Watson Explorer,” a cognitive technology which can gather information from documents integrally and cross-sectionally, and classify and analyze the information in a sophisticated manner. IBM Watson Explorer has powerful functions to search and analyze texts, associate documents, and abstract ideas with focus on patents. On the other hand, SDK has technical expertise and rich experience in research and handling of chemistry-related patent documents. By making the most of this expertise and experience, SDK and IBM Japan successfully enabled the AI-based System to analyze and associate documents in a way specialized for chemistry-related patent documents, and equipped the System with a user interface that gives high readability to patent documents written in Japanese. Thus the new System supports engineers’ work to comprehend and interpret patent related documents in an effective and efficient manner.
In its medium term business plan “The TOP 2021” which was launched in January 2019, the Showa Denko Group prepared a plan to “strengthen utilization of AI/IoT related technologies,” aiming to strengthen the Group’s business foundation. We will accelerate visualization and formalization of technical know-how and processes the Group accumulated through many years of operation, and utilize them for enhancing our corporate value.
Public Relations Office (Phone: 81-3-5470-3235)