Being a Data Scientist, I scout for datasets and I came across an interesting dataset. It contains all the relevant data about US Supreme Court cases. I always thought law as a boring subject but the dataset piqued my interest.
The database to which this introduction pertains to has 70 years of the Court’s decisions,from its first decision in 1946 to the Court’s most recent decision. This dataset is a result of extraordinary work done by Harold J. Spaeth, Distinguished University Professor, emeritus, Michigan State University
Common sense says that the case outcome is decided based on the arguments presented by the Petitioner and Respondent. The final outcome is also based on the Area of Issue and relevant Laws. Still, I went ahead and implemented machine learning Algorithms and created a predictive model.
Much to my surprise, this model was predicting the Outcome of the Supreme Court Cases with 67.4% accuracy. Does that mean that we can use Machine Learning to predict the Judicial Outcome ? I do not have a definitive answer.
To showcase this predictive model, I built a application around this Dataset. You can use this app to predict the outcome probability of an upcoming Court Case by providing the latest case details. You can also visualise the Supreme Court Dataset against various features of the Case.
If you are interested to know more about this model and the predictive ability, please visit Supreme Predictor
Note: If you are a person with legal domain knowledge, I will be interested in knowing your views about this Predictive model. I will also be interested in any kind of collaboration to improve this model.
Citation: Harold J. Spaeth, Lee Epstein, et al. 2018 Supreme Court Database, Version 2018 Release 2. URL: http://Supremecourtdatabase.org