Testing Multiple Reviews
Instead of testing one review at a time, we can classify multiple reviews simultaneously. This demonstrates how the trained model can be applied to a batch of new customer reviews.
Create Sample Reviews
reviews = [
"Excellent quality product. Highly recommended!",
"Worst purchase ever. Don't buy this.",
"Best product in the world. Buy ten immediately!",
"The delivery was quick and the packaging was neat.",
"Absolutely fantastic. Five stars for everything!"
]Preprocess the Reviews
processed_reviews = []
for review in reviews:
review = clean_text(review)
review = remove_stopwords(review)
review = lemmatize_text(review)
processed_reviews.append(review)Convert to TF-IDF
review_vectors = tfidf.transform(processed_reviews)
Predict the Labels
predictions = model.predict(review_vectors)
Display the Predictions
results = pd.DataFrame({
"Review": reviews,
"Prediction": predictions
})
results["Prediction"] = results["Prediction"].replace({
0: "Genuine",
1: "Fake"
})
results
Fake Review Detection System using Machine Learning
VA









