• Thanasekhar B
• Arthi Ramachandran
• Christy K T

Abstract : Sentiment analysis or opinion mining is the computational study of people’s opinions, sentiments, attitudes, and emotions expressed by user-generated content. The essence of sentiment analysis is the text classification task, and different words have different contextual information for classification. In the current sentiment analysis studies, word representation is mostly used. However, the semantic information of a word alone is considered in word representations but ignores the sentiment information of the word. In the proposed sentiment analysis method, sentence representation is used. Two approaches are proposed for sentiment analysis of the full context information which describes the real disaster is input into DistilBERT sentiment classifier and BiLSTM sentiment analysis method to capture the sentiment scores of the context effectively. In the case of large text, using only words would be very tedious and we would be limited by the information extracted from the word embeddings. Hence a word embedding technique in RNN machine learning model BiLSTM is compared with the sentence embedding technique DistilBERT. The evaluation metrics used in the comparison of performance is F1 score. The model acquires accuracy of 84% for BERT along with BiLSTM and 90% for BERT along with DistilBERT. Thus BERT with DistilBERT gives higher performance than BERT with BiLSTM word embedding method.

Opinion Mining, Sentiment Analysis, DistilBert, Word Embedding, Tweet Classifier

"REAL TWEET OPINION MINING BASED ON DistilBERT", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 4, page no.c177-c189, April-2022, Available :http://www.jetir.org/papers/JETIR2204225.pdf

"REAL TWEET OPINION MINING BASED ON DistilBERT", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 4, page no. ppc177-c189, April-2022, Available at : http://www.jetir.org/papers/JETIR2204225.pdf