Sentiment Analysis of Youtube Video Comments Using Artificial Intelligence: The Case of Trump-Zelensky Meeting

Sefer Kalaman

Makale detayları

Sentiment Analysis of Youtube Video Comments Using Artificial Intelligence: The Case of Trump-Zelensky Meeting

Yazarlar:

Sefer Kalaman

Yayıncı:

Istanbul Ticaret University

Ana Başlıklar:

İletişim ve Medya Politikası

Dergi:

İstanbul Ticaret Üniversitesi Sosyal Bilimler Dergisi Vol.24 Issue 52, pp.356-379

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001352529

pISSN:

1303-5495

Anahtar kelimeler:

Özet:

 The diplomatic meeting between Donald Trump and Volodymyr Zelensky on February 28, 2025 had important consequences for international relations and was widely covered both in other countries and in the international press. Especially on social networking sites, many posts were made on the subject, and people commented on the shared content and expressed their feelings and thoughts. From this point of view, the aim of this study is to understand peoples emotional orientations and to determine their perceptions of the meeting by analyzing the emotions of the comments made on YouTube videos of the meeting. In the study, text mining method was used to analyze the comments. 226.000 comments on 4 videos shared on the YouTube platform were collected with the YouTube API interface using Python programming language. Then, the process of cleaning the obtained data was carried out. The cleaned data was divided into tokens and lemmatization was applied and the words were converted into root forms. A sentiment dictionary was then developed to prepare the data for analysis. VADER, Gensim, Pandas, Requests and NLTK libraries were used for data extraction, processing and analysis. Sentiment analysis technique was used to analyze the data obtained. In this context, the ChatGPT-4.5 artificial intelligence model of OpenAI was preferred and sentiment classification was performed with an unsupervised machine learning approach. The data were obtained from 4 videos of the Trump-Zelensky meeting on the YouTube accounts of CNN and Fox News television channels between February 28, 2025 and March 1, 2025. According to the results obtained from the research, anger, neutral and anticipation emotions were the most common emotions in the comments. However, negative emotional orientation predominates in the comments. The most frequently used words in the comments on the posts were Trump, Zelensky-Zelenskyy, America-American, Ukraine and War.

 The diplomatic meeting between Donald Trump and Volodymyr Zelensky on February 28, 2025 had important consequences for international relations and was widely covered both in other countries and in the international press. Especially on social networking sites, many posts were made on the subject, and people commented on the shared content and expressed their feelings and thoughts. From this point of view, the aim of this study is to understand peoples emotional orientations and to determine their perceptions of the meeting by analyzing the emotions of the comments made on YouTube videos of the meeting. In the study, text mining method was used to analyze the comments. 226.000 comments on 4 videos shared on the YouTube platform were collected with the YouTube API interface using Python programming language. Then, the process of cleaning the obtained data was carried out. The cleaned data was divided into tokens and lemmatization was applied and the words were converted into root forms. A sentiment dictionary was then developed to prepare the data for analysis. VADER, Gensim, Pandas, Requests and NLTK libraries were used for data extraction, processing and analysis. Sentiment analysis technique was used to analyze the data obtained. In this context, the ChatGPT-4.5 artificial intelligence model of OpenAI was preferred and sentiment classification was performed with an unsupervised machine learning approach. The data were obtained from 4 videos of the Trump-Zelensky meeting on the YouTube accounts of CNN and Fox News television channels between February 28, 2025 and March 1, 2025. According to the results obtained from the research, anger, neutral and anticipation emotions were the most common emotions in the comments. However, negative emotional orientation predominates in the comments. The most frequently used words in the comments on the posts were Trump, Zelensky-Zelenskyy, America-American, Ukraine and War.

 The diplomatic meeting between Donald Trump and Volodymyr Zelensky on February 28, 2025 had important consequences for international relations and was widely covered both in other countries and in the international press. Especially on social networking sites, many posts were made on the subject, and people commented on the shared content and expressed their feelings and thoughts. From this point of view, the aim of this study is to understand peoples emotional orientations and to determine their perceptions of the meeting by analyzing the emotions of the comments made on YouTube videos of the meeting. In the study, text mining method was used to analyze the comments. 226.000 comments on 4 videos shared on the YouTube platform were collected with the YouTube API interface using Python programming language. Then, the process of cleaning the obtained data was carried out. The cleaned data was divided into tokens and lemmatization was applied and the words were converted into root forms. A sentiment dictionary was then developed to prepare the data for analysis. VADER, Gensim, Pandas, Requests and NLTK libraries were used for data extraction, processing and analysis. Sentiment analysis technique was used to analyze the data obtained. In this context, the ChatGPT-4.5 artificial intelligence model of OpenAI was preferred and sentiment classification was performed with an unsupervised machine learning approach. The data were obtained from 4 videos of the Trump-Zelensky meeting on the YouTube accounts of CNN and Fox News television channels between February 28, 2025 and March 1, 2025. According to the results obtained from the research, anger, neutral and anticipation emotions were the most common emotions in the comments. However, negative emotional orientation predominates in the comments. The most frequently used words in the comments on the posts were Trump, Zelensky-Zelenskyy, America-American, Ukraine and War.

 Donald Trump ile Volodymyr Zelensky arasında 28 Şubat 2025te gerçekleştirilen diplomatik görüşme, uluslararası ilişkiler açısından önemli sonuçlar doğurmuş ve hem diğer ülkelerde hem de uluslararası basında geniş yankı uyandırmıştır. Özellikle sosyal paylaşım ağlarında konu ile ilgili birçok paylaşım yapılmış, insanlar da paylaşılan içeriklere yorum yaparak duygu ve düşüncelerini belirtmişlerdir. Bu noktadan hareketle araştırmada, görüşmeye ait YouTube videolarına yapılan yorumların duygu analizini yaparak insanların duygu yönelimlerini anlamak ve görüşmeye ait algılarını tespit etmek amaçlanmıştır. Araştırmada, yorumları incelemek için metin madenciliği yöntemi kullanılmıştır. YouTube platformunda paylaşılan 4 videoya yapılan 226.000 yorum, Python programlama dili kullanılarak YouTube API ara yüzü ile toplanmıştır. Ardından elde edilen verilerin temizlenmesi işlemi gerçekleştirilmiştir. Temizlenen veriler, tokenlere ayrılarak lemmatizasyon uygulanmış ve kelimeler kök formlarına dönüştürülmüştür. Sonra da eldeki verileri analize hazırlamak için bir duygu sözlüğü geliştirilmiştir. Veri çekme, işleme ve analiz aşamasında VADER, Gensim, Pandas, Requests ve NLTK kütüphanelerinden yararlanılmıştır. Elde edilen verilerin analizini yapmak için duygu analizi tekniği kullanılmıştır. Bu kapsamda, OpenAI firmasına ait ChatGPT-4.5 yapay zeka modeli tercih edilmiş ve duygu sınıflandırması denetimsiz makine öğrenmesi yaklaşımı ile gerçekleştirilmiştir. Veriler, 28 Şubat 2025 - 1 Mart 2025 tarihleri arasında CNN ve Fox News televizyon kanallarının YouTube hesaplarında yer alan ve Trump-Zelensky görüşmesine ait olan 4 videodan elde edilmiştir. Araştırmadan elde edilen sonuçlara göre, yorumlarda en fazla öfke, nötr ve beklenti duyguları yer almaktadır. Bununla birlikte yorumlarda en fazla negatif duygu yönelimi ağır basmaktadır. Paylaşımlara yapılan yorumlarda en çok kullanılan kelimeler ise, Trump, Zelensky-Zelenskyy, America-American, Ukraine ve War olmuştur.

Makale detayları

Sentiment Analysis of Youtube Video Comments Using Artificial Intelligence: The Case of Trump-Zelensky Meeting

Yazarlar:

Sefer Kalaman

Yayıncı:

Istanbul Ticaret University

Ana Başlıklar:

İletişim ve Medya Politikası

Dergi:

İstanbul Ticaret Üniversitesi Sosyal Bilimler Dergisi Vol.24 Issue 52, pp.356-379

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001352529

pISSN:

1303-5495

Artificial intelligence

text mining

sentiment analysis

YouTubeYapay zeka

metin madenciliği

duygu analizi

YouTube

Sayı

Makale detayları

Yazarlar: Sefer Kalaman

Yayıncı: Istanbul Ticaret University

Ana Başlıklar: İletişim ve Medya Politikası

Dergi: İstanbul Ticaret Üniversitesi Sosyal Bilimler Dergisi Vol.24 Issue 52, pp.356-379

Belge Tipi: Makale

Yayın Tarihi: 6/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001352529

pISSN: 1303-5495

Sayı

Yazarlar:

Sefer Kalaman

Yayıncı:

Istanbul Ticaret University

Ana Başlıklar:

İletişim ve Medya Politikası

Dergi:

İstanbul Ticaret Üniversitesi Sosyal Bilimler Dergisi Vol.24 Issue 52, pp.356-379

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001352529

pISSN:

1303-5495