Analyzing Fuzzy Phrases for Emotion Detection Using Distance Based Approach
With the increasing popularity of social media platforms such as Twitter and Facebook the amount of text data available is humungous. Extracting opinions and polarity from this text data available is a hot topic for research work. Many researchers have de
- PDF / 541,177 Bytes
- 9 Pages / 439.37 x 666.142 pts Page_size
- 94 Downloads / 237 Views
, Nikhil Yadav(&) , Nikhil Sharma and Raveesh Garg
,
Delhi Technological University, New Delhi, India [email protected], [email protected], [email protected], [email protected]
Abstract. With the increasing popularity of social media platforms such as Twitter and Facebook the amount of text data available is humungous. Extracting opinions and polarity from this text data available is a hot topic for research work. Many researchers have developed different methods to tackle the problem of sentiment analysis but the main focus by and large has always been on classifying the general tweets. Meanwhile, a large number of fuzzy phrases that assert significant impact on a sentence’s polarity is left untouched. Thus, solving these fuzzy phrases to get membership values for various emotions effectively could widen the scope for sentiment analysis. In this study, we are using one of the most trending topics in India i.e., CAA, NPR and NRC. We are classifying tweets into different emotions to analyze public outlook and then uses the fuzzy sentiment phrases concept to analyze the strength of emotions based on the distance of intensifiers and diminishers from the fundamental words. In our experiment, we got a precision of 0.815 for strong sentiment and 0.713 for weak sentiment, thus showing our method work very well on FSPs. Keywords: Fuzzy sentiment phrases Sentiment analysis CAA NRC NPR
1 Introduction The increasing text data available online and advancement in computational technologies have aided Natural Language Processing (NLP) research to a great extent. The number of social profiles proliferates each day and people express their thoughts, opinions and knowledge on social networking sites. Keeping this in mind, opinion mining becomes an important field of research. One of the most popular social networking sites which serves as one of the major sources for text data is Twitter. The number of Twitter users have grown exponentially with nearly 500 million tweets posted every day. There are 330 million active users and 145 million daily users and a total of 1.3 billion accounts have been created [1]. Twitter Sentiment Analysis (TSA) is a method used in many research papers in order to mine opinions from text data. Sentiment Analysis is basically the process of calculating and analysing the intentions of twitter users about a topic of interest. Sentiment generally refers to emotions, attitudes, feelings or opinions of the users. TSA helps various organizations to gather the emotions and opinions of Twitter users © Springer Nature Singapore Pte Ltd. 2020 C. Badica et al. (Eds.): ICICCT 2020, CCIS 1170, pp. 172–180, 2020. https://doi.org/10.1007/978-981-15-9671-1_14
Analyzing Fuzzy Phrases for Emotion Detection
173
and its analysis. A plethora of sentiment analysis tools can be found online that classify the text to a certain category. However, these tools focus on the basic level of sentiment, that is, they generally classify the text into positive or negative (sometimes also as neutral)
Data Loading...
