1 Introduction

Introduced the concept of register and emphasized its language use and functional significance in specific contexts. The importance of multidimensional analysis in domain research was pointed out, while also highlighting the neglect of semantic analysis in existing research. The aim of this study is to comprehensively understand domain construction and communicative function implementation by incorporating lexical semantics into multidimensional analysis.

Multidimensional analysis method

Multidimensional analysis is influenced by sociolinguistics and multivariate statistical analysis, and identifies high-frequency co-occurrence language features through frequency statistics and factor analysis, forming functional dimensions to distinguish language domains. Biber (1988) extended language features to 67 and clustered them into 7 functional dimensions for comparison of spoken and written language and language categories. Subsequent research may follow these dimensions or explore new dimensions. Multidimensional analysis has been applied in areas such as synchronic comparison of language domains, diachronic evolution, author/translator style, second language writing, and technical tools. But there is a problem of excessive reliance on vocabulary grammar and neglect of semantics. Xiao (2009) proposed multidimensional analysis incorporating semantics, using Wmatrix tool to analyze English variants and registers, marking 141 language features and parsing 9 functional dimensions, effectively distinguishing English variants. Sun Ya and Cui Zixuan (2020) explore the functional dimensions of register based on metaphorical use by combining Wmatrix semantic coding and Biber grammatical features. The aim of this study is to design a multidimensional analysis method with higher degrees of freedom and stronger repeatability, in order to understand semantic intervention and expand the application of technology in language domains.

3 Research Design

The research corpus comes from the Beijing Olympics news from China Daily from 2001 to 2022, with a total of 2577 texts and 975092 symbols. The research steps include extracting language features, descriptive statistics, and factor analysis. Use Stanford Tagger and UCREL Semantic Analysis System tools to automatically identify and annotate lexical grammar and semantic features. Merge semantic features and extract 104 language features combined with grammatical features, totaling 171 language features. Perform factor analysis using SPSS and define factors by clipping factors and their variables through scree plots and load coefficients.

4 Results and Discussion

Through gravel plot analysis, 8 factors were identified for further research. Factor 1 represents "emotional interaction and information presentation", which includes 26 language features, including 23 positive features such as emotional attitude, interpersonal interaction, and information ambiguity, and 3 negative features such as nouns, prepositional phrases, and conjunctions. Factor 2 is "descriptive and abstract narrative of location landscape", which includes 11 language features, 6 positive features, and 5 negative features. Factor 3 is' rights stance 'and contains 7 positive features. Factor 4 is "subjective evaluative and objective factual", which includes 9 language features, 7 positive features, and 2 negative features. Factor 5 is' humanistic and artistic concern ', which only contains 7 positive semantic features. Factor 6 is defined as "health and safety concerns" and includes six positive semantic features. Factor 7 is' facility tools and group activity concerns', which includes 8 language features and 8 positive features. Factor 8 is' technology investment concern ', which includes 5 positive semantic features. These factors reflect the multidimensional characteristics of Beijing Olympics' external propaganda discourse, including emotional interaction, location landscape, rights stance, subjective evaluation, humanities and arts, health and safety, facilities and tools, group activities, and technological investment.

5 Conclusion

This study uses multidimensional analysis to reveal the linguistic characteristics of Beijing Olympics' external news reporting, identifying eight functional dimensions, including emotional interaction, location landscape description, rights stance, subjective evaluation, humanistic and artistic concerns, health and safety concerns, facility tools and group activity concerns, and technology investment concerns. The study has expanded Biber's vocabulary and grammar to the semantic level, overcoming the shortcomings of previous analyses and revealing more refined dimensions of register function and unique communicative functions. The research results demonstrate the core concept and friendly spirit of the Beijing Olympics, but there are limitations, such as not screening vocabulary based on corpus characteristics, unbalanced grammar and semantic features, and the need to further distinguish the potential impact of different levels of semantics in the future.