CSRM-framework for generating actionable knowledge for Social Security Schemes with a special focus on Ayushman Bharat

Abstract

Data Mining applications to Social Security Schemes (SSS) have been one of the most interesting research areas in the recent past. In general, the benefits of SSS are availed by the beneficiaries at different geographical locations at different points in time, thus generating sequential patterns of interest for the stakeholders such as Government bodies or financial institutions to make effective decisions. Typically, SSS launched in the domains of the health sector involves temporal data, and the research in this domain is termed social security data mining (SSDM), where techniques such as sequential pattern mining, sequential rule mining, and association rule mining are in vogue. In this regard, we have proposed a novel data mining framework called the Combined Sequential Rule Mining framework (CSRM-Framework) which is effective in bringing out the actionable knowledge through the activity sequences pertaining to the beneficiaries. The proposed framework was implemented on Ayushman Bharat-Pradahn Manthri Jan Arogya Yojana (AB-PMJAY), a flagship social security scheme launched by the Government of India. We have also proposed a new interesting measure namely Combined Cumulative Lift (CCL) which has the property of estimating the ‘Interestingness’ effectively when activity sequences are combined with characteristic beneficiary data in the context of AB-PMJAY.