Text Mining Related to Halal Implementation in UU Cipta Kerja and UU Halal Product Guarantee No 33/2014

Abstract

This study aims to create a sentiment analysis model for the Halal Product Guarantee Act (JPH) No 33/2014 and the Job Creation Law (UU Cipta Kerja) related to JPH to help facilitate content analysis of existing legislation. The automated sentiment analysis process uses a deep learning approach with the LSTM (Long Short Term Memory) algorithm. The results show that the content of the JPH Law and the UU Cipta Kerja related to JPH contains negative (44.5%), positive (45.0%), and neutral (10.5%) content and is dominated by the words halal certificates, laws, halal products, and BPJPH (halal regulator). The results of the LSTM model test show accuracy in the range of 75% and epoch in the field of 80.