Papaioannou Alexios Serderidis Konstantinos
Sentiment analysis is contextual text mining, identifying and extracting subjective information available in source material. It is one the of the most common text classification tools to support a business or an organization to understand the social sentiment of their product or service based on customers feedback. In this respect, sentiment analysis is nowadays a modern approach to identify customer satisfaction (or dissatisfaction) and facilitate decision making for stakeholders in their effort to improve their business. This study aims at addressing sentiment analysis, in particular airlines sentiment analysis, via Natural Language Processing (NLP) algorithms in the context of customer satisfaction. Baseline algorithms such as Logistic regression, Support Vector Machines, Gradient Boost are explored to establish a reference basis, whereas more elaborate approaches involving neural networks such as Convolutional and Recurrent Neural Networks will be the focus.