Research on the Influencing Factors of Cultivating Interdisciplinary Talents by College English in the Context of Free Trade Port Based on Knowledge Graph

This paper proposes a named entity recognition method based on multitask pointer networks, which uses Transformer and BERT to extract features, thus greatly increasing the fault tolerance of the model. Sequence annotation is realized through the conditional random field model, which can recognize complex nested vocabularies, therefore establishing a knowledge graph related to the Free Trade Port (FTP), constructing the framework of influencing factors for the cultivation of interdisciplinary talents of non-English majors in higher education and preparing a questionnaire on influencing factors. The results show that after 2019 and 2020, the number of articles issued and cited about FTP has demonstrated significant growth. The keywords related to the research hotspots of FTP were clustered, and 14 effective clusters were obtained. The intention to cultivate interdisciplinary talents of non-English majors is positively correlated with students’ knowledge, students’ personality traits, and the economic environment of FTP (P<0.01), with Pearson coefficients of 0.534, 0.287, and 0.335, respectively. To cultivate interdisciplinary talents, the pedagogical principle should be developed, which involves enhancing curriculum, increasing practical experiences, and innovating instructional methods.