Most data analyzed with Ai is structured. This means that the data has been labeled, categorized, and characterized. However, Natural Language Processing helps computers understand unstructuctured text like tweets.
Sentiment Analysis is a part of Nautral Language Processing and use it to determine the mood and emotions of unstructured text. Sentiment analysis allows computers to process a piece of text and then assign numerical values to it for classification and understanding.
Example: Amazon can use sentiment analysis to get a general reaction and rating for a new product based on the comments of users.
Textblob is a python library for Natural Language Processing (NLP), that identifies two data points of sentiment in text. - which labels a text on a range from negative to positive and which labels a text from opinionated to factual.
-- Polarity is a number between 1 - 200 with 100 being neutral. Polarizing text is characterized as negative emotion. This could be anger, dislike, sadness, or frustration. The lower the score, the more polarized the text.
-- Subjectivity is a number between 1 - 100. For example, if a person tweets information about how the blockchain works, this would fall on the higher end of the spectrum. On the otherhand, if a person tweets how much they loved Ethereum, this would be seen as more of an opinion and therefore lower on the scale.