Papers by Khaled Aounallah
Online product reviews are a fundamental part of the decision making process for customers in e-c... more Online product reviews are a fundamental part of the decision making process for customers in e-commerce. These reviews can be classified as positive/negative based on the way they describe the product/venue. They are very influential in making a product successful/unsuccessful by leading people to choose the wrong product. Our project tackles the problem of spam/fake reviews by developing a model, which could classify a given review as either fake or genuine, thereby helping to make more meaningful review information available to the customers. We worked with 6 different models Multinomial Naive Bayes, Logistic Regression, Neural Networks, CNNs, Gradient Tree Boosting and BER. BERT has the highest accuracy of 75%.
Efficient Simulation and Analysis of the Effects of Permeability on the In-Situ Combustion of Heavy Oils
SPE Annual Technical Conference and Exhibition
Uploads
Papers by Khaled Aounallah