IMPLEMENTASI ALGORITMA NAIVE BAYES CLASSIFIER DALAM MENGKLASIFIKASIKAN PUTUSAN MAHKAMAH AGUNG TENTANG KORUPSI
Total View This Week0
Total View This Week0
Institusion
Universitas Stikubank
Author
Fatmawati, Sama
Subject
QA76 Computer software
Datestamp
2020-05-04 03:02:20
Abstract :
Kasus korupsi yang terjadi di Indonesia faktanya setiap tahunnya bertambah. Penelitian akan melakukan klasifikasi putusan korupsi berdasarkan argumen yang sudah dikemukakan saat persidangan. Suatu putusan korupsi dikategorikan pada kelas yang terbukti dan tidak terbukti menggunakan algoritma Naïve Bayes Classifier. Data yang digunakan sebesar 50 data putusan korupsi yang diambil dari website resmi Mahkamah Agung. Data putusan melalui tahap preprocessing meliputi proses Case Folding, Normalisasi Fitur, Tokenizing dan Stopword Removal. Dari hasil uji coba yang dilakukan menggunakan 25 data training dan 25 data testing menghasilkan akurasi sebesar 92%. Hasil klasifikasi divisualisasikan dengan shiny secara online. Shiny menampilkan tabel data, tabel data test, histogram, wordcloud, dan confusion matrix. The fact of corruption cases in Indonesia always increases every years. The researcher will do the classification decision of corruption based on the argument which has already put forward when the court session. A corruption decision categorized to the class which proven or not proven using naïve bayes classifier algorithm. The data that used is as big as 50 data of corruption decision which is taken from the official website account of Supreme Court of Justice. The decision data through the preprocessing stages including case folding process, feature normalization, tokenizing and stopword removal. The result from the testing which is conducted by use 25 training data and 25 testing data has the accuracy result as big as 92%. The result of the classification visualized by online with shiny. Shiny show the data table, test data table, histogram, wordcloud, and confusing matrix.
Kasus korupsi yang terjadi di Indonesia faktanya setiap tahunnya bertambah. Penelitian akan melakukan klasifikasi putusan korupsi berdasarkan argumen yang sudah dikemukakan saat persidangan. Suatu putusan korupsi dikategorikan pada kelas yang terbukti dan tidak terbukti menggunakan algoritma Naïve Bayes Classifier. Data yang digunakan sebesar 50 data putusan korupsi yang diambil dari website resmi Mahkamah Agung. Data putusan melalui tahap preprocessing meliputi proses Case Folding, Normalisasi Fitur, Tokenizing dan Stopword Removal. Dari hasil uji coba yang dilakukan menggunakan 25 data training dan 25 data testing menghasilkan akurasi sebesar 92%. Hasil klasifikasi divisualisasikan dengan shiny secara online. Shiny menampilkan tabel data, tabel data test, histogram, wordcloud, dan confusion matrix. The fact of corruption cases in Indonesia always increases every years. The researcher will do the classification decision of corruption based on the argument which has already put forward when the court session. A corruption decision categorized to the class which proven or not proven using naïve bayes classifier algorithm. The data that used is as big as 50 data of corruption decision which is taken from the official website account of Supreme Court of Justice. The decision data through the preprocessing stages including case folding process, feature normalization, tokenizing and stopword removal. The result from the testing which is conducted by use 25 training data and 25 testing data has the accuracy result as big as 92%. The result of the classification visualized by online with shiny. Shiny show the data table, test data table, histogram, wordcloud, and confusing matrix.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
