QA-FastPerson: Extending Video Platform Search Capabilities by Creating Summary Videos in Response to User Queries

Kazuki Kawamura, Jun Rekimoto

Abstract: In the rapidly evolving field of digital education, the need for efficient and targeted access to information within video content has become critical. This study presents a system designed to enhance the search capabilities of video platforms by generating summary videos that answer user queries. The system uses machine learning and natural language processing techniques to understand complex user queries, pinpoint the exact video segment that provides the answer, and answer user queries more efficiently by providing the user with a summary video around that segment. Preliminary evaluations have demonstrated the system’s potential to accurately identify relevant content and generate effective summaries.