Advanced Technological Strategy Produced by Patent Document Analysis Applying AI Technology

Abstract

This presentation introduces an approach of patent document analysis and an example applying the approach. The approach is constructed of text mining and two AI technology: PLSA and Bayesian Network. This study analyzes about 30,000 patent document data related to wind and air. Applying PLSA can reduce patent documents to dozens of topics, which is useful to understand the overviews simply. Data aggregation and visualization based on the topics enables to understand technical trends and analyze competitor's technical position leading to the strategy of alliance and competition. Modeling the relationships between usage and technology by Bayesian Network gives awareness of new usage of technology owned by companies.