Emb(race): Judicial reform. From traffic stops and arrests to sentencing and parole decisions, use technology to better analyze real-world data, provide insights and make recommendations that will drive racial equality and reform across criminal justice and public safety.
Apache License 2.0
15
stars
12
forks
source link
Compare and Comply for identify poor/inappropriate Sentencing judges and sentences #23
Brief description of your idea: This is NOT AI enabled sentencing
Use/reconfigure existing compare and comply or other NLP/NLC capabilities to provide judges, clerks, attorneys and others the ability to rapidly understand appropriate sentencing guidelines, sentencing histories of similar crimes, sentencing history of the local judiciary and judge to provide cognitive guidance on inappropriate sentencing. Use similar big data capabilities to apply to parole hearings to understand historical guidance and specific parole board history (based on race and other factors) for parole outcomes. Enable all participants to make accurate assessments of the appropriateness of a sentence (or even of a judge before going to trial)
What makes your idea unique?:
Application of known art to a specific use case. COO has not been performed but a brief google search does find similar tools used to apply sentencing based on defendant's history - this would be debiased data to apply sentencing logic to the judge, not the defendant
What would be the impact of your idea if implemented?:
This would be able to provide defendant, attorneys, judges and advocates the ability to rapidly assess possible outcomes as well as provide convincing arguments for most fair/appropriate sentencing. Most importantly this would enable attorneys to immediately address sentencing misconduct and raise objections.
Skills to contribute (e.g. development, architecture, research, design or anything else):
Repurpose of existing compare and comply API stack/cloud architecture
retraining of NLU for specific use case/application
Theme: Judicial Sentencing Reform
Brief description of your idea: This is NOT AI enabled sentencing Use/reconfigure existing compare and comply or other NLP/NLC capabilities to provide judges, clerks, attorneys and others the ability to rapidly understand appropriate sentencing guidelines, sentencing histories of similar crimes, sentencing history of the local judiciary and judge to provide cognitive guidance on inappropriate sentencing. Use similar big data capabilities to apply to parole hearings to understand historical guidance and specific parole board history (based on race and other factors) for parole outcomes. Enable all participants to make accurate assessments of the appropriateness of a sentence (or even of a judge before going to trial)
What makes your idea unique?: Application of known art to a specific use case. COO has not been performed but a brief google search does find similar tools used to apply sentencing based on defendant's history - this would be debiased data to apply sentencing logic to the judge, not the defendant
What would be the impact of your idea if implemented?: This would be able to provide defendant, attorneys, judges and advocates the ability to rapidly assess possible outcomes as well as provide convincing arguments for most fair/appropriate sentencing. Most importantly this would enable attorneys to immediately address sentencing misconduct and raise objections.
Skills to contribute (e.g. development, architecture, research, design or anything else):
Repurpose of existing compare and comply API stack/cloud architecture retraining of NLU for specific use case/application