Incorporating Textual Analysis for Genre Measurement of Narrative Discourse in Business English

In recent years, business English narrative genre research has been characterized by ambiguous concepts, insufficient theoretical basis, and single-genre features, which have weakened the reliability of research conclusions. Accordingly, this paper examines the representative genre of business English narrative, “management discussion and analysis report”, and explores its genre characteristics and its mapping relationship with corporate performance. The measurement framework is based on the discourse level, genre level, professional practice level, and professional culture level and builds a textual analysis path. In order to observe the lexico-grammatical features highlighted in the four levels of discourse, this paper designs the linguistic feature extraction process and the keyword information extraction method involved in the process. This paper improves the LDA model by introducing the emotional information factor, and improves the keyword extraction algorithm in the model based on information theory. According to the designed method for extracting linguistic features, the representative genre text is analyzed in terms of intonation. Through empirical analysis, it is found that the positive tone and negative tone of the retrospective part of the text of the representative genre are correlated with the current year’s performance of the enterprise at a significant level of 1%, and the positive tone is positively correlated with the current year’s performance, while the negative tone is negatively correlated with the current year’s performance. The positive and negative tone of the outlook texts is not significantly related to the performance of enterprises in the next year. The association of genre characteristics with the current year’s performance of the enterprise is evident but not with the next year’s performance. This paper’s text analysis path can be explored in detail for business English narrative discourse genre measures.