The use of machine learning and other algorithmic approaches to provide understanding of how a human conveys emotion for a particular text, image, or audio data in terms of negative, neutral, or positive tone. Sentiment analysis can be considered an aspect of Natural Language Understanding (NLU) in that it aims to gain a machine understanding of human speech, although it can use much simpler statistical methods for basic sentiment classification, and so is usually considered to be a classification application versus a Natural Language Processing (NLP) applications.