COMPARATIVE PERFORMANCE OF RANDOM FOREST AND SUPPORT VECTOR MACHINE ON SENTIMENT ANALYSIS OF REVIEWS OF INDIAN TOURISM

Main Article Content

Smita Selot, et.al

Abstract

Automatic sentiment extractions from social media reviews is new trend in business analysis. It visualizes and summarizes the sentiments extracted from millions of reviews in set of predefined three classes: positive, negative and neutral. Foundations of automated sentiment analysis lies in Natural language Processing (NLP) and Machine Learning (ML) algorithms. Through this paper we are presenting results of applying two robust supervised machine learningalgorithms on Indian tourism reviews: Random Forest(RF) and Support Vector Machine(SVM) and compare the performance of both on the 11K dataset collected through Tripadvisor.com. It is found that using a limited feature, RF outperforms SVM in terms of accuracy and execution time.

Article Details

Section
Articles