Enabling Explainable Artificial Intelligence To Narrate Patent Scenarios in Devising Science Policy Decisions & Technology Forecasts

Abstract

Machine learning methods, offering unique characteristics for variable identification and it’s importance, are now gradually used for prediction, forecasting and devising important scenarios for integral decision-making. Sequential and dynamic growth in technology can fundamentally become a foremost indicator in measuring technology strength in innovative technology-based development. Prudent and systematic data generation is part of the same transition phase playing an equivalent role in creating evidence analysis and service/product development. Assisted solutions of machine learning provide data-driven prediction at core level areas of medical analysis, fraud detection & disease detection, etc. to predict solutions with explainable intelligence. Assisting solutions with more data-driven explainable solutions, during performance of advanced machine learning has created an advanced methodology of generating white box solutions out of the black box interpretation. Explainable artificial intelligence, which gives superior and solution-based AI is not only providing optimal solutions but also provides an impactful decision-making solution. The approach of explainable artificial intelligence being used here represents and mediate assessment for ML algorithms to predict accurate results. Innovation and development, whereby can be identifies via the various innovation indicators, where patents are one of the important solution providers on way for futuristic technology and innovation solution. Patent expansion potentials have been initially proposed by prediction and forecasting of patents, to be considered in identifying technology trends and improving policy decision-making. Now in this paper standardised use of patent values extracted from Organisation for Economic Cooperation and Development (OECD) database of patents for India have been taken to represent technological mandate.