Sentiment Analysis Model for Movie Reviews

This model classifies movie reviews as positive or negative.

Model Details

Model Type: Logistic Regression with TF-IDF features
Training Dataset: IMDB Movie Reviews (40,000 training samples)
Accuracy: ~88.92% on test set
Vectorizer: TF-IDF with n-grams (1-3)

Usage

from huggingface_hub import hf_hub_download
import joblib

# Download model files
model_path = hf_hub_download(repo_id="21f1000330/sentiment-analysis-movie-reviews", filename="sentiment_model.pkl")
vectorizer_path = hf_hub_download(repo_id="21f1000330/sentiment-analysis-movie-reviews", filename="sentiment_vectorizer.pkl")

# Load model and vectorizer
model = joblib.load(model_path)
vectorizer = joblib.load(vectorizer_path)

# Preprocess text (same as training)
def preprocess_text(text):
    from bs4 import BeautifulSoup
    import re
    text = BeautifulSoup(text, "html.parser").get_text() if text else ""
    text = re.sub(r'[^a-zA-Z0-9\s]', '', text) if text else ""
    return text.lower().strip()

# Make prediction
text = "This movie was amazing! I loved it."
preprocessed = preprocess_text(text)
text_vector = vectorizer.transform([preprocessed])
prediction = model.predict(text_vector)[0]
sentiment_result = "positive" if prediction == 1 else "negative"
print(f"Sentiment: {sentiment_result}")

Files

sentiment_model.pkl: Trained Logistic Regression classifier
sentiment_vectorizer.pkl: TF-IDF vectorizer with vocabulary

Citation

Trained on IMDB Movie Reviews dataset for sentiment classification.

Downloads last month: -