Application of Machine Learning Algorithm for Optimal Model Design for Opinion Extraction

Abstract

Evolution of technology introduced huge amount of web data of users of internet that not only use the available web resources, but also give feedback, that generates helpful information. The large size of user data in terms of opinion, feedback, view & suggestions is accessible on web resources, it is normal trend to explore, organize & analyze the reviews for opting right decision. Analysis of sentiment or opinion extraction is an Information Extraction and Natural Language Processing task that helps in identification the user opinions and reviews in form of positive, neutral or negative quotes &comments associated in the text. Various data-driven techniques to opinion extraction are available that applies the sentiment classification. This study explores the efficient classification techniques utilizing KNN using text extraction for a sentiment dataset and performance evaluated in terms of accuracy. Using data from the tweeter dataset, the study evaluates the performance KNN models.