A trend-based patent alert system for technology watch

Abstract

This paper presents development of a trend-based patent alert system (PAS) to find out current trends in patents on industrial technologies. Patent count data based on number of patents filed in specified duration is considered an indicator to define current trend. PAS retrieves data of associated patent counts using extended markup language (XML). Trend extraction algorithm is developed based on linear regression analysis of patent data. Implementation of P AS on real data of textile technologies extracted from online patent database search spaces detected realistic trends for making policy decision on existing technology upgradation and research planning.