Repository files navigation

This project aims to detectfake product reviews usingNatural Language Processing (NLP) andSupervised Machine Learning algorithms. It leverages text-based features and evaluates multiple classifiers to identify deceptive or spam-like reviews from genuine ones.

• Support Vector Machine (SVM) ✅Best performing model
• Random Forest
• XGBoost
• Naive Bayes
• K-Nearest Neighbors (KNN)
• Decision Tree
• Stochastic Gradient Descent (SGD)

• Python 🐍
• Scikit-learn
• Pandas
• Numpy
• Matplotlib / Seaborn
• Natural Language Toolkit (NLTK)

• Preprocessing of textual data (tokenization, lemmatization, stop-word removal)
• TF-IDF Vectorization
• Comparative analysis of model performance (Accuracy, Precision, Recall, F1 Score, ROC-AUC)
• Real-time user review input testing

1. User inputs a product review
2. Preprocessing is applied (cleaning + TF-IDF)
3. Model predicts whether the review isFake orGenuine

A labeled dataset of product reviews (genuine/fake) was used. Each review includes:

• Review_Text
• Label (1 = Genuine, 0 = Fake)

Note: Dataset is not included due to licensing. Please use public datasets or your own for replication.

• Python 3.8+
• pip (Python package manager)

# Clone the repogit clone https://github.com/yourusername/FakeReviewDetection.gitcd FakeReviewDetection