ABSTRACT
India is known for its many languages and cultures. People speak 22 different languages here. People can use the Google Indic keyboard to convey their feelings about any product, news, incidents, laws, games, and so forth. Individual smartphones, tablets, or computers or laptops are used to access social media in their native tongue. Sentiment analysis (SA) is a difficult task in and of itself, but multilingual SA is even more difficult, as different languages have diverse styles of expression, idioms, etc. The vector representation of the WORD2VEC model Bengali words is crucial in Bengali sentiment classification. Words from the same context are found to be more comparable in the vector space of the WORD2VEC model than other words. This chapter presents a novel way to emotion categorization of Bengali comments using WORD2VEC and the Semantic Web. The accuracy obtained by combining the findings of the correlation between the subjectivity score of the words and the WORD2VEC word co-occurrence score is 93 percent.
