SCC-PG: Human-AI Partnership for Knowledge Management and Transfer in Community Social Services

Identifying and accessing social services is critical in aiding vulnerable individuals and families in their move toward self-sufficiency. Communities rely heavily on social service organizations to not only fulfill service needs but to successfully refer community members in need to appropriate services. This requires that each service provider have an accurate understanding of the larger landscape of service providers, what services they can provide, and to whom, at any point in time. We have identified a number of fundamental challenges that these organizations face with regard to the collection and sharing of relevant information. They include keeping information up-to-date, finding time to capture concrete and tacit knowledge, and identifying key holders of knowledge in the organization, just to name a few. This project will examine the potential for the current advances in AI technology to be applied to help social service organizations become more efficient in support of the most vulnerable members of our communities. It will push the boundaries of understanding human-AI partnerships and will result in algorithmic advances in heterogeneous network modeling, social science advances in understanding the factors behind the uptake of AI technology, and interface advances in understanding the most effective means for social workers to engage with AI to capture, maintain, and leverage information in their everyday work. The challenges of social service providers mirror those seen in large commercial organizations which often refer to them as challenges of knowledge management and transfer (KM/KT). We hypothesize that KM/KT innovations can be successfully applied to the social services sector, drastically improving service in our communities and that the introduction of a KM/KT framework among the distributed social service organizations within a local community will provide a unique opportunity to establish an artificial intelligence (AI) layer to leverage that knowledge and amplify the efforts of local service providers. We will develop a proof-of-concept unified data framework for local social service providers to capture and share relevant service information and documents (excluding any personal client information). Over this framework, we will develop an AI layer capable of summarizing and classifying documents, identifying relevant knowledge holders, extracting important information (e.g. dates, eligibility requirements), and creating a complex heterogeneous network model of the social services landscape – all tasks that when combined with service provider expertise, can enhance the services offered across the local social services landscape. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.