Predictive Map

[christyleos]

A predictive map of the outcomes of your case from start to finish.

Variables include:

• police officers
• judges
• pubic defenders
• campaign contributions

[mackdunham]

So this is kind of possible at the macro level via the bulk data reports that the Harris County district clerk puts out. http://www.hcdistrictclerk.com/Common/e-services/PublicDatasets.aspx Though getting the actual filings to figure out/ predict events like if evidence from an officer ususally gets excuded in front of this judge, etc. is probably not possible without a way to pull and catagorize case data from pleadings in the HCDC's electronic case database.

I don't think that campaign contributions are probably not as relevant as you think to predicting case outcomes, so I would advise dropping that.

I would suggest the inclusion of these variables:

• prior convictions
• an expert's testimony prior exclusions
• what ADA is handling the case
• did the grand jury indict
• historical plea deals made by the DA on similar cases

As an example of an Ai doing similar things for family law cases in Ontario take a look here: https://welcome.thistoo.co/

[christyleos]

https://techforjustice-community.herokuapp.com/ for the slack

[christyleos]

GitHub can be found here: https://github.com/TubmanProject/harris-county-predictive-map