Other methods find words with similar polarity using thesaurus such as WordNet. A word’s synonyms are assumed to have similar polarity, and a word’s antonyms are assumed to have opposite polarity.

University of Notre Dame

The core of semisupervised algorithms of induction of affect lexicons lies in the metric of measuring seed word similarities. Besides embedding cosine, there’re other methods.

Other Semisupervised Induction of Affect Lexicons

An on-going but a helpful book resource about NLP
https://web.stanford.edu/~jurafsky/slp3/

Speech and Language Processing (3rd ed. draft) 

Hatzivassiloglou and McKeown (1997) proposed an algorithm that uses syntactic cues. If two adjectives are frequent conjoined by “and” and rarely conjoined by “but”, they are considered similar. And if they’re more likely to be conjoined by “but”, they are likely to have opposite polarity. 

Learn Before

Related