This paper discusses an efficient algorithm for topic level opinion mining and sentiment analysis of online text reviews by using unsupervised topic model, latent dirichlet allocation (LDA) for topic extraction and sentiment analysis of text reviews. The model accuracy is validated on twitter data by evaluating parameters perplexity and loglikelihood and compared with earlier models.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.