Assistant Professor of Computer Engineering, Faculty of Engineering, University of Kurdistan, Sanandaj, Iran
Abstract: (769 Views)
Abstract
Background: Increasing information on news websites and social media has made sentiment analysis a useful field for finding people's opinions. Sentiment analysis includes creating a smart system for gathering and discovering human ideas, feelings, and feedback. Finding Iranian people’s attitudes to the COVID-19 at the start of epidemics using sentiment analysis algorithms was the main goal of this research.
Methods: The research method is computational and is a newer subfield of natural language processing called sentiment analysis. Data was gathered from coronavirus-related user comments of ISNA and Fars news, two popular Iranian news agencies, from February to May 2020. News item numbers collected in this research was 500, which was also collected according to the news of users' opinions, and the number of users' opinions was also close to 300. To analyze user comments, machine learning techniques and tools such as the decision tree algorithm were used to classify positive and negative emotions. In this way, words with negative and positive emotional loads were used to analyze the opinion of users. Machine learning tools and techniques were used for the analysis of user opinions to categorize feelings into positive and negative. Data gathering and preprocessing were done with Python programming language and its associate libraries.
Results: Negative attitude to COVID-19 at the start of pandemics was 21% in Fars news agency and 17% in ISNA. It means that near to 80 percent of the commenting users were optimistic and careless about the disease.
Conclusion: Maybe one reason for the high epidemics and mortality of COVID-19 in Iran at the beginning refers to the wrong attitude and knowledge of the disease. Unlike other countries in the world, Iranian people have spoken with positive words such as happiness, optimism, ease, and inattention about the disease and its problems
Type of Study:
Research |
Subject:
General