Accelerating Patent Evaluation with Text Mining - Ursula Schneider
Ursula Schneider described how text mining is used in Merck KGaA to speed up evaluating patents. While Merck is active in healthcare and life sciences, Ursula described a use case related to performance materials patents, specifically organic light-emitting diodes (OLEDs).
A typical request to the informatics group would be to locate and assess the relevance of patents in the OLED area pertaining to particular compounds, and much of this searching in the patent full text, cross-referencing with the claims section, and assessing relevance was done manually. With the increased volume of patents and queries, Merck needed a way to accelerate this process, and looked to Linguamatics NLP to assist, especially with the searching and highlighting part of the process. Having seen other Linguamatics NLP output displays, Ursula thought this would be well suited to speeding up the manual relevance assessment.
The outcome is a results display with extra fileds that can be exported to Excel for additional filtering and relevance categorization, and reimported. The display includes automatic text highlights that were previously applied manually, and links for the users to inspect the patent full text when needed, with the same highlights displayed. This has substantially speeded up the processing and relevance assessment of OLED patents.