This project of Indonesia Presidential Election 2019 Tweeter Sentiment Analysis
File Sentiment_Analysis_Pilpres_2019.ipynb (google collab) Dataframe : tweet csv
Twitter became a large forum for public discussion during the 2019 presidential election because there were around 5.7 million tweets per day related to the 2019 presidential election. This relatively large amount of data can be used for various analyses, one of which is sentiment analysis.
Tweet sentiment analysis is important for understanding public opinion towards candidates and related issues. This can help campaign teams, media, researchers and the public. Analyzing sentiment can reveal issues highlighted by the public, perceptions of candidates, support, and the impact of misinformation.
Developing machine learning and deep learning models to classify three sentiments (positive, neutral and negative) from tweets related to the 2019 presidential election in Indonesia.
Uses tweet data along with sentiment labels that have been provided & cannot be changed The models used are limited to Random Forest Classifier (machine learning) and LSTM (deep learning)
Python Nltk Keras Pandas
Random Forest (machine Learning) LSTM (deep learning)