Translate 
Abstract
This paper proposes four principles for digital resource construction for language education and carries out practical video resource construction. We believe that extraction and generation will be the main methods for machine resource construction. For resource extraction, this paper first introduces subtitle-based video resource extraction and example sentence extraction. For resource generation, this paper discusses the present limitations and future ideal working framework of large video generation models. The video extraction process proposed in this paper can be applied for other resource construction purposes, and the Large Language Model (LLM) resource generation framework proposed in this paper can be applied to the generation of other resources. We believe that with the digital transformation of education, a large number of machine-constructed resources will emerge.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call