A Study of Output Vocabulary Knowledge in the English Writing Process

Abstract

Abstract Vocabulary acquisition is pivotal in enhancing English writing proficiency. Effective integration of output vocabulary into written English is essential for improving students’ compositional skills. This study proposes a methodology for extracting vocabulary from English textual materials and subsequently applying it to student writing endeavors. To ensure the integrity and accuracy of the text materials utilized, this research employs a Long Short-Term Memory (LSTM) algorithm to perform a comprehensive spelling check on the English writing corpus prior to vocabulary extraction. Further, this paper adopts the high-frequency word list and Term Frequency-Inverse Document Frequency (TF-IDF) techniques to identify and evaluate the significance of vocabulary within the texts. Key vocabulary that significantly impacts word importance classification is preliminarily identified using the Graph Convolutional Network-K Nearest Neighbor (GCKN) algorithm. These pivotal words, termed ‘key nodes, ’ form the basis for constructing a network within the English texts. Utilizing the message-passing mechanism, information from associated nodes is aggregated at the central node, facilitating the acquisition of output vocabulary. The study findings indicate that students, after learning and applying the acquired vocabulary, demonstrate considerable improvements in their English writing capabilities. They exhibit a broader and more sophisticated use of vocabulary, leading to marked enhancements in their writing performance and overall English proficiency.