• Mishra Pankaj S
• Ashish Chaturvedi
• Mukta Agarwal

Opinions area central driver of human behaviour. folks naturally ask for the opinions of others before creating choices, like shopping for merchandise and services, investing, and selection in elections. This consultation is being more and more done victimization microblogging platforms like Twitter, posts on social media, discussion forums or reviews on sites like TripAdvisor.[1][2] Organisations additionally would like feedback on their product and services in order that resources will be allotted efficiently to find new investment opportunities, to publicise and improve product, and to anticipate issues. Consequently, interest has fully grown in an exceedingly field of study known as sentiment analysis to extract which means from the immense amounts of digital opinion knowledge out there. One key feature of a post (or cluster of posts) that's often desired is whether or not its sentiment polarity is positive, neutral, or negative a few subject. This could be accustomed provides a single sentiment signal, or be mass to offer AN opinion over time [3]. It is important, therefore, that the increasing variety of sentiment analysis tools developed for this purpose classify posts as accurately as doable. the most approaches employed in sentiment analysis are in lexicon-based, data-or corpus-based, or a mix of the 2. reckoning on the algorithm used and therefore the Training data, there will probably be wide variations within the results. For example, unsupervised (lexicon-based) methods can perform better across different subject domains, whereas supervised methods (trained, e.g., on product data), may be better in specialist areas. as. Analysis of posts made by the wider public must deal with slang, sarcasm, abbreviations, misspellings, grammatical aspects (e.g., multiple exclamation marks), demographics, and technology changes. For example, emojis and emoticons, which are increasingly used on smartphones, can be used to clarify, enhance, or sometimes reverse the sentiment of a post. Sentiment analysis tools are offered as complete merchandise, however progressively through APIs as web services. This might probably supply organisations the prospect to match merchandise, choose specialist tools betting on needs, and benefit from on-line lexicons and in progress rule development. Sentiment analysis of short social media messages on microblogging platforms like Twitter or Instagram is of high interest to organisations that progressively need to use social media to review the general public mood additionally to or in place of ancient ways of getting feedback, like surveys and opinion polls. An increasing range of specialist tools, which will rate the sentiment of a post during a microblog, are being offered to organisations as web services to cater for this would like. Analysis of microblogging messages should be ready to handle short messages, varied language use, and specifics like emoticons, emojis, and hashtags. Emoticons and emojis are more and more being employed briefly social media messages and seem to own a significant impact on the sentiment of a tweet and therefore the accuracy of classification. For example, one study [4] suggested that using only the emoticon to rate sentiments could achieve accuracy rates of above 80% [5]. further suggested that emoticon sentiment is likely to be more important than text sentiment and may increase accuracy across subject domains. However, [6, 7], in a limited test, cautiously suggested that there may be classification errors with some sentiment analysis tools in the case in which the emoticon sentiment disagrees with the text sentiment. Details of the approach used in developing commercial web services for sentiment analysis are not always available; and therefore comparing them is difficult. It seems that the effect of emoticons and emojis should be considered

Expediency of Social Media Sentiment Analysis Tools with the Support of Emoticon/Emoji

"Expediency of Social Media Sentiment Analysis Tools with the Support of Emoticon/Emoji", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 1, page no.1100-1104, January-2019, Available :http://www.jetir.org/papers/JETIR1901C52.pdf

"Expediency of Social Media Sentiment Analysis Tools with the Support of Emoticon/Emoji", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 1, page no. pp1100-1104, January-2019, Available at : http://www.jetir.org/papers/JETIR1901C52.pdf