Abstract

Social media is currently a place where massive data is generated continuously. Nowadays, majority of\nthe people share their opinions online. Hence, microblogging websites are rich sources of information which have been\nsuccessfully leveraged for the analysis of sociopragmatic phenomena such as belief, opinion and sentiment in online\ncommunication. However, the unprecedented existence of such massive data acts as a double edged sword, one can\neasily get unreliable information from such sources,and it is a challenge to control the spread of false information either\nmaliciously or even inadvertently. The information seeker is inundated with an influx of data. To cope with this , here we\nintroduce a new method for automatically determining the opinion and to assess the credibility of the information. To\nidentify an opinion (positive or negative) about a review, sentiment analysis is performed using Bing Liu\'s dictionary\nand to improve the accuracy of sentiment classification incrementers, decrementers and negation modifiers are\nconsidered. For assessing credibility a new method , \"Automatic Helpfulness Classification\", is introduced. The effects\nare proven by experiments using a large number of reviews and the accuracy obtained is more compared to existing\nmethods.