IIIT-H at SemEval 2015: Twitter Sentiment Analysis – The Good, the Bad and the Neutral!

Dalmia, Ayushi; Gupta, Manish; Varma, Vasudeva

doi:10.18653/v1/s15-2087

Cited by 15 publications

(8 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our model builds on this and achieves higher accuracy on a much smaller training dataset. Ayushi Dalmia [6] proposed a model with a more involved preprocessing stage, and used features like scores from Bing Lius Opinion Lexicon, and number of positive, negative POS tags. This model achieved considerably high accuracies considering the fact that their features were the not the conventional bagof-words, or any n-grams.…”

Section: Related Workmentioning

confidence: 99%

Efficient Twitter sentiment classification using subjective distant supervision

Sahni

Chandak

Chedeti

et al. 2017

2017 9th International Conference on Communication Systems and Networks (COMSNETS)

View full text Add to dashboard Cite

Abstract-As microblogging services like Twitter are becoming more and more influential in today's globalized world, its facets like sentiment analysis are being extensively studied. We are no longer constrained by our own opinion. Others' opinions and sentiments play a huge role in shaping our perspective. In this paper, we build on previous works on Twitter sentiment analysis using Distant Supervision. The existing approach requires huge computation resource for analyzing large number of tweets. In this paper, we propose techniques to speed up the computation process for sentiment analysis. We use tweet subjectivity to select the right training samples. We also introduce the concept of EFWS (Effective Word Score) of a tweet that is derived from polarity scores of frequently used words, which is an additional heuristic that can be used to speed up the sentiment classification with standard machine learning algorithms. We performed our experiments using 1.6 million tweets. Experimental evaluations show that our proposed technique is more efficient and has higher accuracy compared to previously proposed methods. We achieve overall accuracies of around 80% (EFWS heuristic gives an accuracy around 85%) on a training dataset of 100K tweets, which is half the size of the dataset used for the baseline model. The accuracy of our proposed model is 2-3% higher than the baseline model, and the model effectively trains at twice the speed of the baseline model.

show abstract

Section: Related Workmentioning

confidence: 99%

Efficient Twitter sentiment classification using subjective distant supervision

Sahni

Chandak

Chedeti

et al. 2017

2017 9th International Conference on Communication Systems and Networks (COMSNETS)

View full text Add to dashboard Cite

show abstract

“…It is completely play significant part to determine the classes such as positive, negative and neutral [5,6,11]. In lexicon based approach to extract and handle the sentiments effective manner.…”

Section: Proposed Workmentioning

confidence: 99%

Sentiment Analysis using Mode Measure

Edison¹

2017

ijecs

View full text Add to dashboard Cite

In this era, Sentiment Analysis classification and evaluation is a big challenging problem, because most of the people are posting their feelings through Social Media Networks (SMNs) which are reflected as different opinions. In the case of Twitter a tweet only has a maximum limit of 140 characters. Hence, the users are posting opinions like short text, emoticon etc. These kind of opinions are difficult to classify and therefore pose a great challenges to the researchers. In this paper, the sentiments are classified into three classes like positive, negative and neutral emotions. An algorithm called Senti_Mode has been proposed in this paper which is used to handle the sentiments, detect the polarity and find the mode values. Sentiment frequency and percentage were calculated from the Twitter dataset using discrete mode, from these results the observed mode value is 39281 and the percentage of occurrences were 19.55% which are reflected as the most frequent word.

show abstract

“…Formerly, the collection of sentiments are compared with Bing Liu Lexicon dictionary which signify the classes distinctly which are compared with a SetiWordNet [4] and SemEval [23,24] dictionary. The Zipf's Law is federated and the frequency of the sentiments is measured and ranked.…”

Section: Emoticonsmentioning

confidence: 99%