"Custom_Dataset_Comparitive_Study"

"This dataset presents a large-scale analysis of over 1,400 educational YouTube videos curated from community-driven initiatives. The dataset includes structured metadata fields such as video titles, descriptions, captions, and user-generated summaries, focusing on math-related pedagogical content. We propose a methodology to extract, categorize, and assess the educational value of each entry using natural language processing (NLP) and metadata-driven heuristics. Our analysis highlights trends in content structure, subject focus, and engagement styles used by popular explainers. Furthermore, we provide insights into how educators and content creators can optimize math explainers for broader accessibility and classroom relevance. The resulting dataset and evaluation tools support future work in automated educational content recommendation, instructional design, and explainability metrics for learning-focused video material.Index Terms\u2014Educational technology, video metadata, content analysis, natural language processing, online learning, math exposition."