Sentiment Analysis on Movie Reviews

created by : I Gusti Ngurah Ervan Juli Ardana

---

Description

I made this project for the ITS Internal Satria Data (Big Data Challenge) competition. The project uses machine learning to guess how people feel based on movie reviews. let's look deeper

Workflow

1. Import Libraries

2. Load the Data

3. Pre-process the Data

   In the pre-processing stage, various approaches were employed to enhance accuracy:

   • Lowercasing
   • Removing punctuation
   • Eliminating white spaces
   • Removing numbers
   • Eliminating stop words
   • Tokenizing
   • Stemming

4. Feature Extraction

   Two feature extraction methods were explored: TF*IDF and Ngrams. The accuracy analysis indicated that TF*IDF yielded superior results.

5. Model Development

   Several models were tested, including logistic regression, Naive Bayes, Random Forest, and SVM. Based on accuracy results, logistic regression was chosen as it demonstrated the highest accuracy.

6. Hyperparameter Tuning

   Hyperparameter tuning was performed using gridsearchcv to optimize the logistic regression model's parameters.

7. Test Prediction

   After completing all these steps, the model achieved an accuracy of 0.881.