Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Negative Polarity Levels for Sentiment Analysis


Affiliations
1 Department of Information Systems, CU Cairo, Egypt
     

   Subscribe/Renew Journal


The goal of Sentiment analysis is defining the attitude of a writer with respect to some topics or the overall sentiment polarity of a text, such as positive or negative. Online sentiments have a big effect in making a decision in business. There are several challenges in analyzing and evaluating sentiments. More than 60% of sentiments face a negative polarity challenge. In most research, negative consists of two levels: implicit and explicit. But we present new criteria for analyzing negative sentiments. The criteria include five negative levels that can effect on the word or sentence polarity. This paper proposes a new technique to improve the accuracy by analyzing negative reviews. We investigate the effect of evaluating negation in sentiment analysis based on word level. We also discuss the negative words and phrases types, with respect the conflict with several expressions. Our experimental results indicate that by evaluating and classifying for negative, precision relative to human ratings increases with 10%.


Keywords

Accuracy, Challenges, Explicit Negative, Implicit Negative, Reviews, Sentiment Analysis.
User
Subscription Login to verify subscription
Notifications
Font Size

Abstract Views: 234

PDF Views: 1




  • Negative Polarity Levels for Sentiment Analysis

Abstract Views: 234  |  PDF Views: 1

Authors

El-Din Doaa Mohey
Department of Information Systems, CU Cairo, Egypt

Abstract


The goal of Sentiment analysis is defining the attitude of a writer with respect to some topics or the overall sentiment polarity of a text, such as positive or negative. Online sentiments have a big effect in making a decision in business. There are several challenges in analyzing and evaluating sentiments. More than 60% of sentiments face a negative polarity challenge. In most research, negative consists of two levels: implicit and explicit. But we present new criteria for analyzing negative sentiments. The criteria include five negative levels that can effect on the word or sentence polarity. This paper proposes a new technique to improve the accuracy by analyzing negative reviews. We investigate the effect of evaluating negation in sentiment analysis based on word level. We also discuss the negative words and phrases types, with respect the conflict with several expressions. Our experimental results indicate that by evaluating and classifying for negative, precision relative to human ratings increases with 10%.


Keywords


Accuracy, Challenges, Explicit Negative, Implicit Negative, Reviews, Sentiment Analysis.