Twitter Sentiment Analysis Using Syntactic Action Rule-Based Decision Regression

Abstract

            Sentiment Analysis (SA) is the most effective technique to determine people's sentiments or emotionsand the text's attitude regarding any event.SA comes with the rapid development of the Twitter microblogging service users can post their messages (tweets) to followers (friends) to find analyzing people's feelings and emotions. According to the sentiment analysis is a classification problem of the attitude of the text of the tweet.Due to increasing tweets' increasing from users' opinions, text detection sentiment is a low accuracy process.The proposed Syntactic Action Rule-based Decision Regression (SARDR) algorithm is done to classify the tweets and extract the actionable patterns based on SA of Twitter to solve this problem.Initially collect the tweets dataset then pre-processing done to remove unwanted data and noise then feature extraction for collect the relevant text data information calculate the verb weightages, finally trained into SARDR algorithm prepared to classify the sentiment of text as positive, negative or neutral.The proposed algorithm is the corpus-based Twitter sentiment analysis generated based on the structure of words and the verbs' correct form.The experimental results are shown to improve the classification accuracy compared with previous algorithms.