Multiclass-Subjectivity Feature Clustering-Based Twitter Sentiment Analysis Using Content-Based Deep Feature Neural Network for COVID-19 Wave Opinions

Resumen

Recent day’s social media is big discussion area for content discussion on social media have sentiment facts depending on the user opinions because of different user have different opinion to share in twitter tags. Due to improper analysis in content analysis, the subjectivity of original information be varied due to opinion projection of words are trajectory. So, the prediction of the trajectory contents failed to produce classification in accuracy in the sense of precision rate. To resolve this problem, we propose a Multiclass-Subjectivity Feature Clustering (MSFC) Based Twitter Sentiment Analysis (SA) Using Content Based Deep Featured Neural Network (DFNN) for Covid-19 Wave Opinions to find the positive, negative opinions. This estimates the Hashtag Content Indexing Rate (HTCIR) depends on the user tweet tags accumulation carried out tagging speech of words. Then the tweet term sentiment was observed by lexical analysis using Lexical Subject Impact Score (LSIC). From the sentiment score lexical features point are averaged into mean depth threshold values and considering features are grouped into clustering using MSFC. Then the cluster be ranked using Opinion Ranking Weight (ORW). Once trained as a DFNN, it can classify the terms and select features based on the references. This proposed system produces higher efficiency compared to the other system as well in SA attains high calcification and precision accuracy than other methods.