A new visual navigation system for exploring biomedical Open Educational Resource (OER) videos.

Area of Science:

• Biomedical Sciences
• Educational Technology

Background:

• The proliferation of biomedical open educational resources (OER) videos presents challenges for effective content retrieval.
• Existing keyword- and content-based search methods are insufficient for navigating large biomedical video libraries.

Purpose of the Study:

• To develop a novel visual navigation system for efficient information seeking from a massive corpus of biomedical OER videos.
• To provide users with interactive visual and textual clues for semantically rich and user-friendly content exploration.

Main Methods:

• Collected and processed approximately 25,000 biomedical science YouTube videos (4000 hours total).
• Extracted semantic clues from audio, visual, and textual elements of each video.
• Stored clues in a metadata database indexed by a high-performance text search engine.
• Developed a prototype system rendering dynamic web pages for interactive video exploration.

Main Results:

• A prototype system is publicly accessible, demonstrating functional effectiveness.
• User study results indicate the system is user-friendly and effective for information seeking.
• The system facilitates efficient content exploration among massive biomedical OER videos.

Conclusions:

• Users can efficiently find relevant videos and pinpoint specific information segments.
• The system allows intuitive and convenient previewing of essential video content.
• This tool significantly improves the accessibility and usability of biomedical OER videos.