A Comparative Sentiment Analysis Of Sentence Embedding Using Machine Learning Techniques

A Comparative Sentiment Analysis Of Sentence Embedding Using Machine Learning Techniques

Table of Contents




Abstract

Analyzing sentiment is a process to identify the opinion of a text. It is also known as opinion mining or emotion Artificial Intelligence (AI). People post comments in social media mentioning their experience about an event and are also interested to know if the majority of other people had a positive or negative experience of the same event. This classification can be achieved using Sentiment Analysis. Sentiment analysis takes unstructured text comments about product reviews, an event, etc., from all comments posted by different users and classifies the comments into different categories as either positive or negative or neutral opinion. This is also known as polarity classification. Sentimental analysis can be performed by Text analysis and computational linguistics. This work aims at comparing the performance of different machine learning algorithms in performing sentiment analysis of Twitter data. The proposed method uses term frequency to find the sentiment polarity of the sentence. The performance of Multinomial Naive Bayes, SVM, and Logistic regression algorithms in sentence classification were compared. From the results, it is inferred that logistic regression has achieved the greatest accuracy when it is used with the n-gram and bigram model.

 

  • Author Keywords

    • Sentiment Analysis,
    • sentence classification,
    • performance,
    • SVM,
    • Multinomial Naive Bayes,
    • Logistic regression
  • IEEE Keywords

    • Sentiment analysis,
    • Support vector machines,
    • Machine learning algorithms,
    • Logistics,
    • Twitter,
    • Machine learning,
    • Classification algorithms

 

Introduction

Analyzing sentiment is a process to find out the opinion of a text. People post comments on social media mentioning their experience about an event and are also interested to know if the majority of other people had a positive or negative experience of the same event. Analyzing sentiment is a process of knowing users’ emotions for a particular item which may be an event or topic or individual of recent trends. Sentiment analysis can be done at three levels and they are sentence, aspect, and document level. Twitter is a highly rich source of information for deciding the quality of any product. The Twitter platform uses tweets which is in sentence form to denote opinions. Therefore, sentiment analysis at the sentence level is used for examining sentiments. Sentiment analysis over Twitter provides the companies with an effective way to find the opinions of people towards their newly launched products. The goal is to calculate the sentiment accuracy of sentences that were extracted from the text of tweets. Sentiment analysis can be done for twitter data which will classify the tweet as either positive or negative. This analysis helps concerned organizations to find opinions of people about their product, events, so on from the tweets. The opinion words are the most challenging part of sentiment analysis. An opinion word may be positive or negative depending on the situation. The meaning of the content will not be altered by traditional text processing systems when there is a little change in the words. But sentiment analysis can change the meaning of the content when there are changes in two words. For example, “The phone is ringing” is different from “the phone is not ringing”. The processing is done sentence by sentence. The informal sentence on twitter can be understood by the user whereas the system cannot understand. This work compares the performance of SVM, Multinomial Naive Bayes, and Logistic regression in sentence classification of twitter data. This paper is organized in such a way that the literature survey is discussed in section II, followed by the description of proposed work in section III, the results of sentiment analysis are discussed in section IV and the conclusion in section V

Conclusion

Sentiment analysis is a process to identify the opinion of a text. People post comments on social media mentioning their experience about an event and are also interested to know if the majority of other people had a positive or negative experience of the same event. The goal is to calculate the sentiment accuracy of sentences that were extracted from the text of tweets. The sentiment analysis of the tweet helps to find whether the sentiment of the tweet on particular products, events, etc., is positive or negative. The most challenging task in sentiment analysis is to identify an opinion word as either positive or negative. By changing the parameter of the machine learning algorithms is possible to get the best accuracy. This work compared the performance of three machine learning approaches includes SVM, Multinomial Naïve Bayes, and Logistic regression. The Logistic Regression has achieved an accuracy of approximately 86% when the bigram model was used. Logistic regression performs better when compared to other supervised machine learning algorithms for Twitter sentiment analysis. Future work can be of analyzing the fluctuation in the performance of the sentiment analysis algorithm when multiple features are considered. A further active learning technique like expected error reduction, pool-based sampling, uncertainty sampling, and so on shall be utilized to detect Twitter sentiments and to increase the confidence of decision-makers.

About KSRA

The Kavian Scientific Research Association (KSRA) is a non-profit research organization to provide research / educational services in December 2013. The members of the community had formed a virtual group on the Viber social network. The core of the Kavian Scientific Association was formed with these members as founders. These individuals, led by Professor Siavosh Kaviani, decided to launch a scientific / research association with an emphasis on education.

KSRA research association, as a non-profit research firm, is committed to providing research services in the field of knowledge. The main beneficiaries of this association are public or private knowledge-based companies, students, researchers, researchers, professors, universities, and industrial and semi-industrial centers around the world.

Our main services Based on Education for all Spectrum people in the world. We want to make an integration between researches and educations. We believe education is the main right of Human beings. So our services should be concentrated on inclusive education.

The KSRA team partners with local under-served communities around the world to improve the access to and quality of knowledge based on education, amplify and augment learning programs where they exist, and create new opportunities for e-learning where traditional education systems are lacking or non-existent.

FULL Paper PDF file:

A Comparative Sentiment Analysis Of Sentence Embedding Using Machine Learning Techniques

Bibliography

author

P. A and K. S. Priya,

Year

2020

Title

A Comparative Sentiment Analysis Of Sentence Embedding Using Machine Learning Techniques

Publish in

2020 6th International Conference on Advanced Computing and Communication Systems (ICACCS), Coimbatore, India, 2020, pp. 493-496,

Doi

10.1109/ICACCS48705.2020.9074312.

PDF reference and original file: Click here

 

+ posts

Somayeh Nosrati was born in 1982 in Tehran. She holds a Master's degree in artificial intelligence from Khatam University of Tehran.

Website | + posts

Professor Siavosh Kaviani was born in 1961 in Tehran. He had a professorship. He holds a Ph.D. in Software Engineering from the QL University of Software Development Methodology and an honorary Ph.D. from the University of Chelsea.

Website | + posts

Nasim Gazerani was born in 1983 in Arak. She holds a Master's degree in Software Engineering from UM University of Malaysia.