Robust Extreme Learning Machine based Sentiment Analysis and Classification

In recent times, S entimentalAnalysis (SA) acquires important attention in the process of decision making, primarily implied for the classification and extraction of the sentiments exist in the online reviews posted by the user. SA could be assumed as a sentiment classification (SC)issue where the online reviews experiences classification into negative and positive polarities based on the words available in the online reviews. This study focuses on the design of Robust Extreme Learning Machine Based Sentiment Analysis and Classification (RLM-SAC) model. The presented REIM-SAC model majorly aims to determine the nature of sentiments exist in the input data. Primarily, the input data is thoroughly pre-processed to get rid of unwanted data, which helps in enhancing the classification accuracy and minimizing the computational complexity. In addition, the presented REIM-SAC model applies ELM model to allocate proper class labels to it. To adjust the parameters of the ELM model, comprehensive learning particle swarm optimization (CLPSO) technique was used. The performance assessment of the RELM-SAC model is experimented with using benchmark database and the outcomes are scrutinized under numerous aspects. The simulation outcomes pointed out that the RELM-SAC method has obtained improved outcomes than other models.