Identifying and communicating routes using graph networks

A device, comprising:

• one or more memories;
• and
• one or more processors, communicatively coupled to the one or more memories, configured to:
• receive, from one or more data sources, geospatial data identifying roads and buildings in a geographical region;
• transform the geospatial data identifying the roads and the buildings into geospatial objects that correspond to the roads and the buildings;
• correlate shelter data that identifies shelters in the geographical region with the geospatial objects to generate correlated data that includes the geospatial objects and the shelter data;
• process the correlated data to add bounding boxes around the geospatial objects corresponding to the shelters in the geographical region, to generate modified correlated data;
• correlate road weight and type data with the modified correlated data to generate final geospatial data for the geographical region;
• convert the final geospatial data into a directed graph for the geographical region;
• receive, from one or more data sources, event data identifying an event and inaccessibility of roads, caused by the event, in the geographical region;
• update the directed graph for the geographical region based on the event data and to generate an updated directed graph;
• receive, from a client device of a user, location data identifying a location of the client device in the geographical region;
• process the updated directed graph and the location data, with a machine learning model, to identify one or more shelters associated with the location of the client device;
• and
• perform one or more actions based on identifying the one or more shelters.

A non-transitory computer-readable medium storing instructions, the instructions comprising:

• one or more instructions that, when executed by one or more processors, cause the one or more processors to:
• transform geospatial data, identifying roads and buildings in a geographical region, into geospatial objects that correspond to the roads and the buildings;
• correlate shelter data that identifies shelters in the geographical region and road weight and type data with the geospatial objects to generate final geospatial data for the geographical region;
• convert the final geospatial data into a directed graph for the geographical region;
• receive event data identifying an event and inaccessible roads, caused by the event, in the geographical region;
• update the directed graph for the geographical region based on the event data and to generate an updated directed graph;
• receive, from a client device of a user, location data identifying a location of the client device in a geographical region;
• process the updated directed graph and the location data, with a machine learning model, to identify one or more shelters associated with the location of the client device;
• and
• perform one or more actions based on identifying the one or more shelters.