ABSTRACT
In this section we introduce you to sentiment analysis and guide you through two distinct approaches. Firstly, we teach you how to apply a lexical approach for assigning sentiment polarities and classes to text documents. Secondly, we show you how to employ a deep neural network for sentiment analysis. In regards to the latter, we show you how to leverage a pretrained BERT architecture for sentiment classification, explain its structure, and training procedure, before we teach you how to fine-tune a BERT model to assign custom sentiment values, and cover how to evaluate classifiers based on deep neural network algorithms.
