ABSTRACT
People frequently seek out the advice of others and ask for recommendations before making decisions. Customers often search for a product before purchasing it, and online reviews have a major impact on their decision. Deep learning employs artificial neural networks that are meant to mimic how people think and learn as opposed to traditional machine learning algorithms, which use simpler rules and mathematical models. Some of the deep learning methods include Deep feedforward neural network, Convolutional Neural Network (CNN), Recurrent Neural Network, Long Short-term Memory and Gated Recurrent Unit (GRU). The goal of cloud computing is to deliver hosted services through the internet. CNN’s generated local features, and the long-term dependency mechanism of GRU achieved excellent aspect extraction results, resulting in better sentiment classification of aspects within hotel and car datasets used for evaluation. The service repository component locates various cloud services and serves as a source of available cloud services for service consumers to use.
