Arabic Restaurant Reviews Emotion Classification Using NLP

Built an NLP-based deep learning model to classify Arabic restaurant reviews into six emotions: joy, anger, sadness, disgust, surprise, and neutral. I performed Arabic text preprocessing and normalization, extracted features using TF-IDF with unigram and bigram representations, and developed a neural network classification model using Keras. The project achieved approximately 94% validation accuracy and demonstrates practical experience in Arabic NLP, text preprocessing, feature extraction, and emotion classification.