Twitter User Sentiment Analysis for RUU Omnibuslaw Using Convolutional Neural Network

EasyChair Preprint 5333

Abstract

The general function of social media is for online interaction with many people. Moreover, social media have functions for sharing information, discussion, and giving an opinion media about some topics that a lot of people talk about, one of that media is Twitter. An atopic will show many opinions and different responses from everyone. This study was for making an analysis opinion from social media Twitter user about Rancangan Undang-Undang Omnibuslaw topic using a Convolutional Neural Network method wich one of Deep Learning method. This study has been done a sentiment analysis with opinion data from many different people through the tweet they making, Preprocessing and weighting are done using Word2vec which give 84% result accuracy of an algorithm from 10-time testing. Based on 2.820 tweet data, the result is 1.320 data of positive sentiment, and 1.500 data of negative response for the Rancangan Undang-Undang Omnibuslaw topic in Indonesia.

Keyphrases: Convolutional Neural Network (CNN), Sentiment Analysis, Twitter

@booklet{EasyChair:5333,
  author    = {Popon Dauni and Muhammad Ali Ramdhani and Dimas Ramdhani Suryapratama and Wildan Budiawan Zulfikar and Jumadi},
  title     = {Twitter User Sentiment Analysis for RUU Omnibuslaw Using Convolutional Neural Network},
  howpublished = {EasyChair Preprint 5333},
  year      = {EasyChair, 2021}}