A post-pandemic world resulted in economic upheaval, particularly for the cities' communities. While significant work in NLP4PI focuses on national and international events, there is a gap in bringing such state-of-the-art methods into the community development field. In order to help with community development, we must learn about the communities we develop. To that end, we propose the task of community learning as a computational task of extracting natural language data about the community, transforming and loading it into a suitable knowledge graph structure for further downstream applications. We study two particular cases of homelessness and education in showing the visualization capabilities of a knowledge graph, and also discuss other usefulness such a model can provide.