Rahmadini, Rina (2025) Pemodelan Prediksi Hasil Putusan Pengadilan Berdasarkan Data Posita Perkara Perceraian Di Pengadilan Agama Kota Padang Sidempuan Menggunakan Pendekatan Machine Learning. Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6032231236-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6032231236-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy |
Abstract
Penelitian ini bertujuan membangun model prediksi hasil putusan perkara perceraian di Pengadilan Agama Kota Padang Sidempuan dengan pendekatan machine learning berbasis data posita. Metode yang digunakan mencakup preprocessing teks posita, ekstraksi alasan perceraian berbasis kata kunci, normalisasi dan encoding variabel, serta pelatihan model klasifikasi biner. Enam algoritma diterapkan, yakni Decision Tree, Naïve Bayes, K-Nearest Neighbors (KNN), Random Forest, LightGBM, dan XGBoost. Evaluasi model dilakukan menggunakan metrik akurasi, precision, recall, F1-score, F2-score, dan AUC, dengan mempertimbangkan karakteristik data yang bersifat imbalanced dan biner. LightGBM dan XGBoost memberikan performa terbaik dengan keseimbangan antar metrik dan nilai AUC tertinggi, masing-masing 0,80 dan 0,81. Analisis feature importance menunjukkan bahwa hasil mediasi merupakan fitur paling berpengaruh dalam proses klasifikasi hasil putusan. Meski terdapat indikasi overfitting pada beberapa algoritma, penelitian ini memberikan kontribusi awal bagi pengembangan sistem pendukung keputusan di lingkungan peradilan. Implikasi manajerialnya meliputi pengembangan sistem informasi perkara yang lebih adaptif, serta potensi integrasi fitur rekomendasi input alasan perceraian ke dalam aplikasi pengadilan seperti SIPP. Dengan demikian, hasil penelitian ini menjadi pijakan awal dalam mempersiapkan institusi peradilan untuk integrasi sistem prediksi putusan berbasis kecerdasan buatan di masa mendatang.
====================================================================================================================================
This study aims to develop a predictive model for court verdicts in divorce cases at the Religious Court of Padang Sidempuan using a machine learning approach based on posita (legal claim statement) data. The methods include text preprocessing of posita, keyword-based extraction of divorce reasons, normalization and encoding of variables, and binary classification model training. Six algorithms were implemented: Decision Tree, Naïve Bayes, K-Nearest Neighbors (KNN), Random Forest, LightGBM, and XGBoost. Model evaluation was conducted using accuracy, precision, recall, F1-score, F2-score, and AUC metrics, considering the imbalanced and binary nature of the data. LightGBM and XGBoost achieved the best performance, demonstrating balanced metric scores and the highest AUC values of 0.80 and 0.81, respectively. Feature importance analysis revealed that the outcome of mediation was the most influential feature in classifying verdicts. Despite indications of overfitting in some models, this study offers an initial contribution toward the development of decision-support systems in the judicial environment. The managerial implications include the enhancement of a more adaptive case information system and the potential integration of a recommendation feature for divorce reason inputs into court applications such as SIPP. These findings serve as a foundational step in preparing judicial institutions for the integration of AI-based verdict prediction systems in the future.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Ketidakseimbangan Data, Machine Learning, Perkara Perceraian, Posita Perkara, Prediksi Putusan, Divorce Cases, Imbalanced Data, Machine Learning, Posita, Verdict Prediction |
| Subjects: | T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.84 Heuristic algorithms. T Technology > T Technology (General) > T58.62 Decision support systems |
| Divisions: | Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT) |
| Depositing User: | Rina Rahmadini |
| Date Deposited: | 30 Jul 2025 10:19 |
| Last Modified: | 23 Sep 2025 07:42 |
| URI: | http://repository.its.ac.id/id/eprint/124170 |
Actions (login required)
 |
View Item |