1. Introduction

The importance and challenges of academic English writing in international academic exchanges were discussed, and it was pointed out that Chinese scholars face special difficulties due to language differences between Chinese and English. In academic English writing, it is difficult to grasp the characteristics of academic style, and vocabulary research involves subject coverage, functional applications, and lexical chunks. However, vocabulary knowledge is still a problem that needs to be solved. This study aims to compare the difficulty of academic English vocabulary between Chinese scholars and native English speakers, provide vocabulary usage suggestions for Chinese scholars, and provide inspiration for academic English teaching and vocabulary difficulty theory.

2. Related research

Discussed the concept of vocabulary difficulty style and its application in academic English writing. Vocabulary difficulty is a sub concept of vocabulary richness, reflecting the preference for vocabulary use, and its core elements include the depth and breadth of vocabulary knowledge. Word frequency, word range, word familiarity, and word meaning richness are indicators for measuring vocabulary difficulty. Vocabulary difficulty plays an important role in evaluating writing ability and distinguishing academic paper types. The multidimensional analysis method provides multiple perspectives for understanding vocabulary, and achieves dimensionality reduction and clustering of multiple features through statistical analysis methods, becoming an important paradigm for discourse variation research. This study will use multidimensional analysis to compare the differences in vocabulary difficulty characteristics between Chinese scholars and native English speakers in computational linguistics papers.

3. Dimensionality reduction statistics of vocabulary difficulty features

Introduced the dimensionality reduction statistical process of vocabulary difficulty characteristics in academic English writing by Chinese and English scholars. The researchers used the Semantic Scholar Open Research Corpus to collect 1132 computational linguistics papers published by Chinese scholars and native English speakers between 2010 and 2020, and established a corpus of Chinese English academic papers. Use TAALES 2.2 tool to annotate and count vocabulary difficulty features, which includes 484 features covering 13 categories. Through feature pre screening, non normal distribution and multicollinearity features were removed, resulting in the retention of 150 features. Using a random forest model, select 49 features that have strong ability to distinguish between Chinese and English scholars' papers. Through factor analysis, 49 language features were divided into six dimensions, with a cumulative variance explanation of 61.012%. After factor rotation, there are ultimately 45 valid language features remaining. Calculate the dimension scores of English papers by Chinese and English scholars, and one-way ANOVA shows significant differences in the first five dimension scores, which can distinguish the papers of Chinese and English scholars.

4. Comparative analysis of vocabulary difficulty dimensions

This study explores the comparison of vocabulary difficulty in academic English writing between Chinese and English scholars through multidimensional analysis. Analyzing the co-occurrence, load value, and functional commonality based on language features, five main dimensions were revealed: lexical generality, binary word string stability, ternary word string stability, word processing difficulty, and the correlation between actual and function words. Research has found that native English speakers tend to use binary and ternary word strings with strong universality and stability, as well as content words with high processing difficulty and weak relevance, while Chinese scholars tend to use word strings with weak universality and stability, as well as content words with low processing difficulty and strong relevance. These differences indicate that Chinese scholars should reduce vocabulary difficulty, improve semantic clarity, and enhance their awareness of dialogue with readers in academic writing. At the same time, it is suggested that Chinese scholars use the resource library constructed by native English speakers to assist writing. Teachers can adopt data-driven teaching methods and use the vocabulary information in the resource library to assist in lesson preparation and teaching. In addition, Chinese scholars and English teachers can make full use of online resources such as EAT, USF, and WordNet to better cater to the language style of native English speakers.

5. Conclusion

This chapter summarizes the main findings of a comparative study on the difficulty of academic English writing vocabulary between Chinese and English scholars, pointing out that Chinese scholars have weak universality and stability of vocabulary in their academic papers, low processing difficulty, strong relevance of actual words, and weak relevance of function words. The research results have implications for improving the academic English writing ability and teaching practice of Chinese scholars. It is recommended to use the language resource library of native English speakers to assist writing, combined with multi word chunk teaching, to enhance the flexible application ability of similar words in form, sound, and meaning. At the same time, it is pointed out that quantitative data alone is not enough, and qualitative methods such as interviews need to be combined to explore the causes of differences in vocabulary use and supplement existing corpus analysis.