ANALYZING THE IMPACT OF YOUTUBE ON DISSEMINATION OF CONSTRUCTION ENGINEERING AND MANAGEMENT KNOWLEDGE

Abstract

The exponential growth of digital learning platforms has revolutionized knowledge dissemination across disciplines, yet their efficacy in Construction Engineering and Management (CEM) education remains inadequately explored. This study addresses this research gap by investigating YouTube's role in CEM knowledge dissemination, focusing on three critical objectives: (1) identification of key CEM knowledge areas, (2) quantitative analysis of user engagement sentiment, and (3) thematic exploration of discussions through comment clustering. Key CEM knowledge areas were systematically identified through an exhaustive review of peer-reviewed literature and curriculum analysis of the top 50 universities offering CEM-related courses. "Construction Management" emerged as the primary domain, encompassing 26 related sub-areas for in-depth analysis. A robust mixed methods approach was employed, integrating systematic literature review, sentiment analysis, and K-means clustering. The study analyzed a comprehensive dataset of 316,046 comments and 72,564 replies across 27 CEM subjects, utilizing YouTube's API for data extraction and advanced Python libraries (NLTK, pandas, scikit-learn) for analysis. Sentiment analysis revealed a predominantly positive reception of CEM content on YouTube, with 53.06% of comments and 69.06% of replies expressing positive sentiments. Negative sentiments were limited to 14.30% of comments and 10.17% of replies, while neutral sentiments accounted for 32.69% and 20.77% respectively. Engagement levels varied significantly, with "Construction Process Management," "Construction Equipment and Personnel Management," and "Construction Management" receiving the highest number of comments. Clustering analysis uncovered a vibrant online learning community, with key themes including appreciation for clear explanations, technical discussions on topics such as Building Information Modeling (BIM), and engagement with industry trends. The analysis revealed global participation and cross-cultural professional development, evidenced by clusters in various languages. Study limitations include potential sentiment classification bias due to the complexity of natural language and the specificity of analyzed CEM subjects. Future research avenues include longitudinal analysis of sentiment evolution, cross-platform comparative studies, and integration of machine learning techniques for enhanced thematic extraction. The findings provide empirical evidence supporting YouTube's effectiveness as a knowledge dissemination platform in the CEM domain. These insights have significant implications for CEM educators, content creators, and industry stakeholders, offering data-driven strategies to optimize online learning resources and align content development with learner needs and industry trends.