Patent Application Trend

Patent application trend analysis uses computational methods to understand patterns and predict future trends in patent filings. Current research focuses on applying graph representation learning, dynamic graph models, and natural language processing (NLP) techniques, including transformer-based architectures like BERT, to analyze patent text and relationships between patents, companies, and technologies. This research aims to improve patent classification, predict future patent applications, and assess patent value, ultimately aiding in more efficient innovation management and strategic decision-making for businesses and researchers. The insights gained can also help prevent redundant research and development efforts.