Court Judgement Prediction Using Long-Short Term Memory and Convolutional Neural Network

Abstract

Ever thought about law and enforcement, how much will be the data in the sector of law and enforcement. Even for one single judgement, the judge and the lawyer and related personals must refer several references like previous cases, judgments, victims, situations, duration, time, evidence, count of evidence, judges, advocates, they must study each, and the whole thing ultimately related to their current case. Along with time the historical data in the legal sector also increased a lot, nowadays the industrialists are trying to collect such legal information and data to estimate the judgements by analysing and compiling the collected data with the help of court case rules. Even though they tried a lot to find the judgements, it becomes failure without a proper implementation. But if the machine learning algorithms are implemented for the judgement prediction, it can be observed victorious and accurate prediction as compared to the normal conventional method of forecasting the judgements. The major intention of the investigation is to predict the court judgements with the help of hybrid neural network model (long short-term memory and convolutional neural network). Through optimization of features, the most applicable datums through means of long short-term memory along with the perception of convolutional neural network model is carried out to expect the court judgements.