IIT Kanpur has created an AI-driven system known as the Family Score that aims to assist governments in identifying families in need of welfare support by utilising existing government data. This innovation seeks to diminish the need for extensive, repeated household surveys, which can be both time-consuming and costly. The institute emphasised that the Family Score system could expedite the identification process for eligible households, providing a much-needed solution to welfare delivery issues.

The project emerged from a collaboration between IIT Kanpur and the Government of Andhra Pradesh. Recently, the initial findings were presented to Andhra Pradesh Chief Minister N Chandrababu Naidu and key government officials, illustrating the potential impact of this system on welfare identification.

Functionality and Accuracy of the Family Score System

The Family Score system assesses household socio-economic conditions by analysing pre-existing datasets held by government bodies. The initiative utilises around 15 datasets that contain metrics related to socio-economic factors. AI models were trained using this information and subsequently calibrated against scores produced through a comprehensive statewide survey.

Once the models were sufficiently trained, they began generating Family Scores for households included in the available datasets, eliminating the necessity for new data collection or surveys. IIT Kanpur reported that this ensemble of machine learning models has shown capabilities to exceed traditional surveys in accurately determining the socio-economic conditions of families, requiring less time, effort, and resources.

The next phase of the project will expand the Family Score system across Andhra Pradesh and will be integrated with the AP State Data Centre (AP SDC). This phase will feature a large-scale validation and calibration exercise, further enhancing the system’s credibility and effectiveness.

Coverage and Future Implications of the Project

IIT Kanpur noted that the existing datasets possessed by the government are capable of producing Family Scores for approximately 1.35 crore households out of an estimated 1.72 crore in Andhra Pradesh. This results in an effective coverage of around 75 per cent of households within the state, demonstrating significant potential for the welfare assessment system.

The initiative has been spearheaded by IIT Kanpur’s Wadhwani Centre for Developing Intelligent Systems (WCDIS) and the Wadhwani School of AI & Intelligent Systems (WSAIS), in association with the Airawat Research Foundation (ARF). IIT Kanpur’s Director Professor Manindra Agrawal highlighted that the aim is to harness AI and governmental data to gain insights into families’ socio-economic situations, thereby enabling more targeted welfare distribution.

In addition to the efficiency of the system, Professor Nitin Saxena, Dean of the Wadhwani School of AI & Intelligent Systems, pointed out the interpretability aspect. Each Family Score comes with a rationale, which allows government officials to comprehend the basis of the assessments and retain control over final decisions. This aspect is crucial for ensuring trust in AI-driven methodologies.

Transformative Potential for Welfare Delivery

The traditional method of conducting household surveys is often labour-intensive and requires significant financial investment, particularly when addressing millions of households. The Family Score system proposes a model that leverages existing governmental data rather than relying exclusively on periodic surveys. Should this approach be validated successfully, it could enable governments to refresh socio-economic evaluations more dynamically.

The success of the Family Score initiative will, however, rely heavily on the quality and timeliness of the governmental data employed in generating the scores. The upcoming statewide validation phase is expected to be pivotal in determining the accuracy of the AI system in identifying household needs.

IIT Kanpur has asserted that the primary goal of the project is to enhance welfare delivery without replacing governmental decision-making processes. Officials will continue to hold accountability for final decisions, emphasizing that the initiative aims to unite fragmented data sources and produce evidence-based insights for public welfare policy implementation.

Long or Short, get news the way you like. No ads. No redirections. Download Newspin and Stay Alert, The CSR Journal Mobile app, for fast, crisp, clean updates!

App Store –  https://apps.apple.com/in/app/newspin/id6746449540 

Google Play Store – https://play.google.com/store/apps/details?id=com.inventifweb.newspin&pcampaignid=web_share