Abstract:

The widespread use of Artificial Intelligence (AI) has also been evaded the field οf education, ranging from chatbots to fact checkers or even “writers”. This has sparked a profound debate among students, educators, and the general public. As educational institutions grapple with policy-making and implementation, there is an urgent need for data-driven insights into how these stakeholders perceive the so-called AI revolution. This study addresses this gap by conducting a systematic computational analysis of public sentiment as reflected in social media discourse, and particularly in YouTube, a popular platform for audiovisual content consumption but also informal learning and information acquisition.

For this study, we will perform an analysis of thousands of user comments harvested from five influential YouTube videos regarding AI in Education. The videos will be selected in terms of popularity as appear when we search “AI in Education”. To keep our analysis thorough and scientifically sound, we will implement a methodology that utilizes four language models. More specifically, our framework will integrate the following models for comments analysis at a sentence level: SentimentR for general linguistic polarity, the NRC Lexicon for granular emotional categorization, the RoBERTa-base and SentimentBERT (SBERT).

The primary objective of this research is to identify the Emotional Dissonance that often appears in the educational sector's digital transition. By utilizing different language models, the study aims to move beyond simple positive/negative polarity but it seeks, to quantify specific psychological triggers, such as Fear, Anticipation, and Trust, to determine if any possible public anxiety is rooted in technical limitations or in the perceived replacement of human mentorship and qualifications.

This research constitutes a clear and adaptable methodology that will help educational researchers and administrators to move beyond anecdotal evidence. By measuring the emotional barriers to AI adoption, our research provides a practical path for a more Human-Centered AI integration. We conclude that the future of educational management lies in Analytical Curation, where sentiment data is used to validate institutional strategies, ensuring that technology serves to aid, rather than replace, the human foundations of learning.

Keywords:

AI in Education, Sentiment Analysis, Public Perception.