Design and Implementation of Knowledge Extraction Tool for Teaching Video Based on Verbal Behavior Characteristics

The rapid advancement of artificial intelligence technology, tightly intertwined with the field of education, highlights the vast potential for leveraging AI in the analysis and assessment of teaching behavior. Verbal behaviors in teaching, as the primary way of knowledge transmission, has a direct and profound impact on teaching quality. However, existing researches in this domain focus on classroom environments and assessment of teaching effectiveness, and relatively little research has been conducted on the extraction of knowledge from teaching verbal behaviors. Researchers typically identify knowledge in teaching videos through manual observation and recording, a time-consuming and labor-intensive process that makes it challenging to extract knowledge from the extensive trove of teaching videos. In this paper, a knowledge extraction tool based on teaching video verbal behaviors was designed and implemented. It provides knowledge searching, positioning, and extracting functions, with a view to provide help for teaching researchers to analyze knowledge from the perspective of teaching verbal behavior. Additionally, it greatly improves the efficiency of extracting knowledge from teaching videos.