From Material Generation to Vocabulary Acquisition: A New Model of Second Language Vocabulary Acquisition Driven by Large Language Model

Abstract

Along with the advancement of globalisation, foreign language proficiency has become a key skill, of which vocabulary acquisition is the core of language learning. The article aims to explore the effectiveness of Large Language Models (LLMs) generative materials in second language vocabulary acquisition and examine their effectiveness among learners of different language levels. Based on the comprehensible input hypothesis and the attention hypothesis, the study conducted a two-week vocabulary acquisition experiment with Japanese majors at a university in Guangdong Province, using experimental research method and quantitative analysis. The experiment compared the differences in vocabulary acquisition effects between learning materials generated by the Large Language Model and traditional instructional materials, and evaluated the effectiveness of the new vocabulary acquisition model by statistically analysing the scores of 39 students on immediate and delayed tests. The results of the study showed that the learning materials generated by the Large Language Model significantly outperformed the traditional materials in immediate vocabulary acquisition, but there was no significant difference in the effectiveness of the two methods in the delayed test. In addition, learners with different language levels did not show significant differences in vocabulary acquisition when using the learning materials generated by the Large Language Models.