Care-Network for Reintegration

Description

As a growing city with a poverty rate twice the national average, South Bend residents are in need of easy ways to find and access social services in the city. A particularly vulnerable population is those who have fulfilled their criminal sentences and find themselves in need of assistance for reintegration back into society. This is a complicated and resource-starved process that requires expert knowledge and familiarity with local services. However, the current system is tedious, confusing, and its usage requires trained "navigators" to find resources for clients. To alleviate these issues, in this project we aim to:

• Integrate some of the expert knowledge through the use of SOTA NLP language models embeddings, and Information Network Modeling to provide an easy-to-use semantic search for social services via queries and relevant recommendations during the search process
• Target the aforementioned population taking into account their particular conditions such as job, health and other needs, criminal history, location, past referrals and more
• Explore the use of auxiliary qualitative data such as Social Determinants of Health, referral notes, and Needs Trees to augment our results and recommendations