I am establishing an absent sidewalk prioritization for bond dollar expenditures. Each missing sidewalk segment will be scored according to various criteria including proximity to businesses, parks, jobs, etc.
I could perform a simple Euclidean buffer, but I believe the analysis would be more sophisticated using a network analysis. In other words, to find a way to score the sidewalks according to the destinations captured within x-distance walk-shed along the street network.
In what I have seen done, and in all of the research and tutorials I've found, network analysis is done using point data as the input rather than lines. So, for a service area, a business (point) is used as the input and analyzed according to the network (streets).
Because absent sidewalks are line segments - not points - I am not sure if what I would like to do is even possible.