Predicting the Category and the Length of Punishment in Indonesian Courts Based on Previous Court Decision Documents

Abstract

Among the sources of legal considerations are judges’ previous decisions regarding similar cases that are archived in court decision documents. However, due to the increasing number of court decision documents, it is difficult to find relevant information, such as the category and the length of punishment for similar legal cases. This study presents predictions of first-level judicial decisions by utilizing a collection of Indonesian court decision documents. We propose using multi-level learning, namely, CNN+attention, using decision document sections as features to predict the category and the length of punishment in Indonesian courts. Our results demonstrate that the decision document sections that strongly affected the accuracy of the prediction model were prosecution history, facts, legal facts, and legal considerations. The prediction of the punishment category shows that the CNN+attention model achieved better accuracy than other deep learning models, such as CNN, LSTM, BiLSTM, LSTM+attention, and BiLSTM+attention, by up to 28.18%. The superiority of the CNN+attention model is also shown to predict the punishment length, with the best result being achieved using the ‘year’ time unit.

We Know You Are Living in Bali: Location Prediction of Twitter Users Using BERT Language Model

“…In addition, several other features could be incorporated, such as time zone information [46] and friend networks [43,46]. In addition, deep learning methods using attention mechanism may also be explored to combine features directly in a single model [74].…”

Criminal Court Judgment Prediction System Built on Modified BERT Models

“…3, the steps conducted in this research include dataset collection and parsing, pre-training the BERT model followed by fine-tuning it and doing a model performance comparison and evaluation. The dataset used in this study comes from the Indo-Law dataset [16]. The Indo-Law dataset contains data on court decision documents from the website of the Supreme Court of the Republic of Indonesia (known as Mahkamah Agung Republik Indonesia) that meet the following requirements:…”

“…However, previous studies were conducted for different purposes. Nuranti et al [16] predict the categories and length of punishment in firstlevel court using multi-level learning (CNN+attention). Nuranti et al [17] observed the effectiveness of several machine learning and deep learning methods to recognize 10 legal entities in Indonesian court decision documents.…”

Investigation of Lawsuit Process Duration: A Machine Learning and Process Mining Approach

“…The selected articles mainly belong to three categories. The first one concerns the prediction of outcomes from judgments [17][18][19][20] and classification of legal documents [21]. The second branch of works investigates the duration of a case in machine learning [16,22,23], and a mathematical [24][25][26][27] perspective.…”

“…Nuranti et al [20] used CNN + attention mechanism to predict first-level judicial decisions in Indonesian courts. They made use of documents regarding the decision of 22,630 cases.…”