Aggregation and Normalization of Geographic Data

[BryceStevenWilley]

Just getting some useful correspondence with @nonprofittechy and @calvinmetcalf out from the depths of my email to be more visible / searchable. Most of these ideas are their's, not mine.

Idea: we want admins of DA servers to know who they are helping (and more importantly, who they aren't so efforts can be focused there), using GIS data. We don't want:

• to just end up with a heatmap
• the resulting graphic to be de-anonymized in anyway (this one is kinda big, might split into another issue)

We do want:

• the graphic to be interesting
• the information to still be usable with a small number of users of the form (~200)
• a simple interface, without a lot of different knobs, maybe a few simpler ones

@calvinmetcalf's suggestions included some of the following:

• Normalization would help with avoiding a heatmap. It would have to be domain specific, i.e. for eviction related topics, normalize based on evictions per 1000 people or evictions per 1000 households in the area. We can see this ERSI demo table per zip for more data.

• Zip codes are generally a bad way to aggregate when they cross municipal boundaries, but in MA they're mostly okay. No good solutions for aggregation between zips and municipalities. Neighborhoods in Boston would be good, but they aren't well defined.