Based on the topic models database and time density of caption's dialogue, designed a adaptive, iterative incremental learning algorithm, which could meanwhile carry out topic detection, words clustering and calculating words' used degree. Lastly, based on words' used degree, by calculating degree of difficulty of caption's dialogue, graded videos' corresponding to captions, constructed a large multimedia database for oral Chinese teaching.
Liu, H. (2013). Constructing of multimedia resources for second language teaching based on intelligent information processing of movie resources. International Journal of Emerging Technologies in Learning, 8(5), 42–46. https://doi.org/10.3991/ijet.v8i5.3045