Sentiment classification

Sentiment in text can be either positive or negative. We will stick to this definition for this chapter. Sentiment mining is an active field of research. A starting point for someone to learn the various aspects of sentiment mining is through the book Sentiment Analysis and Opinion Mining, by Bing Liu.

Broadly speaking, sentiment mining problems are solved using the following techniques:

Dictionary methods

Sentiment lexicons, in which words are categorized as positive and negative, are used in this technique. There are several lexicons available today:

• SocialSent is a domain specific lexicon available from Stanford University: https://nlp.stanford.edu/projects/socialsent/
• Wordnet is a lexical database for English from Princeton: http://wordnet.princeton.edu/
• There are several sentiment annotated datasets available at Bing Liu's website: https://www.cs.uic.edu/~liub/FBS/sentiment-analysis.html

We match these lexicons to the tokens (words) in the input data and we can then...