Open Access Open Access  Restricted Access Subscription Access

A Comprehensive Method for Sentiment Analysis on Twitter Posts

Kanan Joshi, Mr. Dhaval Patel



In recent years, with ascent of users of social media, researchers get drawn to sentiment analysis. The challenge is that sentiment-dependent info from multiple sources does not seem to be thought of usually in existing sentiment classification techniques. Sentiment analysis is incredibly vital as a result of it helps folks and organizations in their decision-making process. There is vast explosion of “sentiments” on the market from social media like Twitter, Facebook and user forums. We have got set to figure with micro-blogging website Twitter as a result of five hundred million tweets get tweeted on daily bases and most characters that are allowed are one hundred forty. This paper explores the evolution and a comprehensive methodology for sentiment analysis. Understanding of the feelings of human massages towards totally different entities and product, permits higher services for discourse advertisements and analysis of market trends. We tend to use machine learning classifiers to detect sentence polarity. Our input to classifier is tweets without Stopwords. Unigram method is used for feature selection. The classifiers we use are Bayesian network, Random Forest, and k-Nearest Neighbor. The differences between classification time of decision Bayesian network, Random forest and KNN are about an order of magnitude, based on accuracy. This study includes a brief introduction describing the area of application of opinion mining, and some definitions useful in the field. The most commonly used methods are mentioned and some alternative ones are described. Experiment results are presented which show that KNN algorithm gives best results than other algorithms.


Keywords: Sentiment analysis, machine learning classifiers, natural language processing


Cite this Article

Kanan Joshi, Dhaval Patel. A Comprehensive Method for Sentiment Analysis on Twitter Posts. Journal of Multimedia Technology & Recent Advancements. 2017; 4(2): 1–5p.

Full Text:



  • There are currently no refbacks.

This site has been shifted to