Contains 2 simple projects which have the same goal, which is to analyze sentiment, whether the sentiment is positive, negative or neutral. The sentiment analyzed is Indonesian sentiment.
Rule-based sentiment analysis here is a sentiment analysis method using predefined logic rules. The rules are determined according to the list of keywords that have been given in the words.txt file whether they are negative or positive words without using data training.
Sentiment analysis Naive Bayes here is a technique for analyzing sentiment using the Naive Bayes machine learning method. This method uses data training to learn the probability of each word appearing in a positive, negative or neutral. Which later from the results of the Naive Bayes model will produce a probability that can later identify a word including positive or negative or neutral words.
To run either of the projects, we have the same steps:
- go to one of the naive_bayes or rule_based projects with
cd "folder_project_name" - create a virtual environment from python first using
python -m venv .venv - activate the virtual environment according to your Operating System
- install all the requirement libraries needed by running
pip install requirements.txt - run the main.py file with
python -m main
Although this program lacks accuracy in solving problems, it was created for my own personal learning and as an experiment in machine learning.