Sentiment Analysis with Natural Language Processing Methods
Moving beyond traditional content analysis methods, the research centers on Natural Language Processing (NLP) techniques. In her study, Merve Boyacı Yıldırım employed advanced analysis methods such as BERT-based sentiment analysis, BERTopic topic modeling, and word co-occurrence networks. Through these methods, thousands of comments on videos containing the keyword "Gaza" were detailed across their emotional, humanitarian, and political layers.
The findings indicate that user reactions are predominantly shaped around negative emotions. In particular, the level of visual violence, tragic stories, and the political stances of news channels significantly influenced the emotional tone in viewer comments. The study presents striking, data-driven results regarding how "emotional publics" are formed in the digital environment.
Bringing a unique perspective to the relationship between framing and reception in conflict reporting, this study has added a valuable resource to the literature for understanding how digital discourse is shaped during times of crisis. We congratulate our faculty member on her academic success and wish her continued success.
You can use the link below to access the full text and citation details of the study.
Article Access:
Academic Profile:
© Copyright 2022 Istanbul Gelisim University All Rights Reserved.
