Sentiment Classification-How to Quantify Public Emotions Using Twitter

Sentiment Classification-How to Quantify Public Emotions Using Twitter

Prerna Mahajan (Institute of Information Technology And Management, New Delhi, India) and Anamika Rana (Shobhit University, Meerut, India)
Copyright: © 2018 |Pages: 15
DOI: 10.4018/IJSKD.2018010104
OnDemand PDF Download:
No Current Special Offers


This article describes how with the tremendous popularity in the usage of social media has led to the explosive growth in unstructured data available on various social networking sites. Sentiment analysis of textual data collected from such platforms has become an important research area. In this article, the sentiment classification approach which employs an emotion detection technique is presented. To identify the emotions this paper uses the NRC lexicon based approach for identifying polarity of emotions. A score is computed to quantify emotions obtained from NRC lexicon approach. The method proposed has been tested on twitter datasets of government policies and reforms, more about current NDA government initiatives in India. The polarity components apply and classify the tweets into eight predefined emotions. This article performs both quantitative and sentiment analysis processes with the objective of analyzing the opinion conveyed to each social content, assign a category (+ve, -ve & neutral) or numbered sentiment score. The assigned scores have been classified using six different machine classification algorithms. Good classification results are achieved with the data.
Article Preview

2. Sentiment Analysis

Now a day’s studies have increasingly focused on semantic analysis as a process that extracts sentiment information that is different from statistical, syntactic or semantic information. Specifically, semantic analysis deals with the extraction of user’s opinion and emotions, extraction over different blogs, discussion forum, social sites etc. people’s interaction and curiosity towards social media and internet, has attracted many researchers to explore the user’s behavior towards a specific topic.

There is different classification of Sentiment analysis: Document level, sentence level, aspect level. Document level sentiment classification deals with the extraction of opinion paring with words from reviews and detecting the polarity of these opinionated terms (Pang et al., 2002; Turney, 2002). Sentence level classification identify a sentence as subjective or objective and hence, also called as subjectivity classification (Wiebe et al., 1999). These subjective sentences are considered as small documents and further classified by extracting and classifying opinion as positive and negative. Aspect/feature level; gives a more fine-grained model, which extracts opinions expressed against different aspect/ features of the entity. This involves extraction of opinions and aspects, and categorizing them into similar classes, determining the polarity of opinions and summarization of results. Figure 1 shows the different classification levels and subtask of aspect level sentiment analysis (Liu, 2012, Liu & Zhang, 2012).

Complete Article List

Search this Journal:
Open Access Articles: Forthcoming
Volume 14: 4 Issues (2022): 2 Released, 2 Forthcoming
Volume 13: 4 Issues (2021)
Volume 12: 4 Issues (2020)
Volume 11: 4 Issues (2019)
Volume 10: 4 Issues (2018)
Volume 9: 4 Issues (2017)
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing