Predictive Analytics of Indian IRS: Using IR Usage Data to Forecast Societal and Policy Impact from the Last Decade (2016–2026)

Abstract

The Indian Institutional Repositories (IRs) have transformed from passive digital archives into proactive information ecosystems over the last decade. The aims of present paper to develop a predictive model for forecasting the societal and policy impact of scholarly research hosted by Indian IRs. This study investigates evolution of Indian IRs via utilizing a 10-year dataset (from 2016 to 2026), and analyses longitudinal usage of information from 125 prestigious Indian IRs, mainly listed in global directories like OpenDOAR and ROAR, including download metrics, metadata growth, and user demographics. The study uses statistics tables, correlation analysis, regression modelling to examine repository growth patterns, and to predict future developments. This research also utilizes time-series forecasting (ARIMA) predictive analytics to analyse usage data to predict how this data will drive policy and its societal impact. This study is a first decade long longitudinal study in context of Indian IRs to predict the 90% accurate non-academic impact of repository stored scholarly outputs. The findings also suggest that the predictive models can assist policymakers in aligning research output with national policy or development goals or in other words with the Viksit Bharat 2047 vision.