Two-Step Model for Sentiment Lexicon Extraction from Twitter Streams
Ilia Chetviorkin, Natalia Loukachevitch · 2014
In this study we explore a novel technique for creation of polarity lexicons from the Twitter streams in Russian and English. With this aim we make preliminary fil-tering of subjective tweets using general domain-independent lexicons in each lan-guage. Then the subjective tweets are used for extraction of domain-specific sen-timent words. Relying on co-occurrence statistics of extracted words in a large un-labeled Twitter collections we utilize the Markov random field framework for the word polarity classification. To evaluate the quality of the obtained sentiment lex-icons they are used for tweet sentiment classification and outperformed previous results. 1