This paper introduces the problem of semi-automatically building decision models from eligibility policies for social services, and presents an initial emerging approach to shorten the route from policy documents to executable, interpretable and standardised decision models using AI, NLP and Knowledge Graphs. There is enormous potential of AI to assist government agencies and policy experts in scaling the production of both human-readable and machine executable policy rules, while improving transparency, interpretability, traceability and accountability of the decision making.
A recent study challenges the common belief that income support programs like SNAP reduce employment, finding that for individuals with a work history, receiving SNAP benefits can actually increase long-term employment.
This research summary presents findings from a randomized controlled trial demonstrating how mRelief’s simplified SNAP application significantly increases application rates among eligible individuals.
This research explores how software engineers are able to work with generative machine learning models. The results explore the benefits of generative code models and the challenges software engineers face when working with their outputs. The authors also argue for the need for intelligent user interfaces that help software engineers effectively work with generative code models.
This article examines how the decentralization of safety net programs after welfare reform has led to growing inequality in benefit generosity and access across U.S. states.
The article analyzes the impacts of Arkansas's Medicaid work requirements, finding that while coverage losses were reversed after the policy was halted, it did not improve employment and led to negative consequences such as increased medical debt and delayed care.
This article explores how legal documents can be treated like software programs, using methods like software testing and mutation analysis to enhance AI-driven statutory analysis, aiding legal decision-making and error detection.
This article analyses ‘digital distortions’ in Rules as Code, which refer to disconnects between regulation and code that arise from interpretive choices in the encoding process.
This paper analyzes the unique challenges of conducting participatory design in large-scale public projects, focusing on stakeholder management, fostering engagement, and integrating participatory methods into institutional transformation.
Recent studies demonstrate that machine learning algorithms can discriminate based on classes like race and gender. This academic study presents an approach to evaluate bias present in automated facial analysis algorithms and datasets.
This paper introduces a method for auditing benefits eligibility screening tools in four steps: 1) generate test households, 2) automatically populate screening questions with household information and retrieve determinations, 3) translate eligibility guidelines into computer code to generate ground truth determinations, and 4) identify conflicting determinations to detect errors.
