Political Messaging through Language Patterns and Sentiment Analysis in Twitter Messages - Focusing on the US President Twitter


Kwak, Myunsun

피인용수 : 0(자료제공 : 네이버학술정보)



This paper makes a text mining analysis of the US President Donald Trump and the former President Barack Obama’s remarks on their respective Twitter accounts to examine what words are mainly used by these powerful politicians, and through that what political messages they attempt to convey. The study analyzes Twitter messages from 2017 to 2019 in terms of language patterns and sentiment analysis. The analysis of the language patterns is focused on the use of parts of speech, the use of capital letters, and exclamation points. The results show that Trump uses pronouns most and verbs and nouns are followed, while Obama uses nouns most and pronouns and verbs are next. Trump uses capital letters much more than Obama and the most frequently used word is GREAT while Obama hardly uses capital letters. In the case of Trump, he uses exclamation points with the word, you most while the word everybody is the case for Obama. As there is no nonverbal language elements such as intonation in text messages, Trump tends to use many capital letters and exclamation points to express his feelings as nonverbal elements of language. Through sentiment analysis it was found that both politicians use more positive words than negative words. However, Trump uses more negative words than Obama. As the messages on Twitter have become a primary means for political communication, the two politicians tend to use more positive words to create the positive images.


1. Introduction
2. Literature Review
2.1 Twitter’s Role in Politics on Social Media Engagement
2.2 Nonverbal Language Elements in Text Based Communication
2.3 Sentiment Analysis in Twitter Messages
3. Methodology
3.1 Data Collection and Classification of Valid Data
3.3 Data Analysis
4. Results and Discussion
4.1 Analysis of Twitter Language Use Patterns
4.2 Analysis of the Use of Capital Letters and Punctuation Points as Linguistic Elements of Twitter Messages
4.3 Sentiment Analysis
5. Conclusion


  • Kwak, Myunsun Daejeon University / Professor


자료제공 : 네이버학술정보

    ※ 기관로그인 시 무료 이용이 가능합니다.

    • 6,600원

    0개의 논문이 장바구니에 담겼습니다.