This exercise aims to create a sentimental analysis that can be used in a variety of text analyses to use variable data.
data used from carant-ai/indonesian_sentiment_datasets obtained from hugging faces with the amount of data of 1.030.393 rows but due to resources limitations then limited to 30,000 rows or 30% of the data
This Sentiment analysis model uses two models - Logistic Regression model - Naive Bayes's model Logisitic regression model looks better in this case
Deployment uses a Logistic Regression model with the extraction of the topic from positive, neutral and negative sentiment so that the user knows what the topic should be concerned about.
This exercise is limited to 30,000 rows and needs to be implemented with more data and using other models as well as need to be enhanced related to lematisation and stemming