An Improved K-Means Clustering with Machine Learning Based Sentiment Analysis and Classification Model

Sentiment analysis (SA) involves the task of automatically extracting the sentiments from user reviews via Natural Language Processing (NLP), data mining, and machine learning (ML) models. A major intention of SA is to identify the opinion and sentiments. It becomes helpful in the decision making of the customers whether to purchase an item or not. This paper develops an improved K-means clustering with random forest (RF) classification model called IKC-RF for effective SA. In order to handle the huge amount of online product reviews, K-means clustering technique is utilized to cluster the sentiments into appropriate class labels. Initially, the online product reviews are preprocessed and feature extraction process takes place. Next, clustering process is carried out by K-means clustering and finally, classification is done by RF technique. The application of clustering technique helps to handle the massive increase in dataset. The performance of the IKC-RF model is evaluated and the results are examined under distinct aspects.