Abstract In recent years, the accuracy and linguistic fluency of translation based on intelligent algorithms have been relatively close to that of human beings, and the application of intelligent algorithms has gradually become a research hotspot in the academic world. In this paper, it is found through relevant research that semantic retention, cultural appropriateness, linguistic fluency, content comprehensiveness and text readability are important factors in the application of intelligent algorithms for the translation of literary works for cross-cultural communication. To explore the degree of influence of these factors on the translation quality of literary works, this paper analyzes them using a multiple linear regression model. The first step in building the mathematical model for multiple linear regression is to construct it. Then, the formulas for calculating the goodness-of-fit and significance test of multiple linear regression equations are introduced. Finally, the regression model was tested using residual analysis and multicollinearity. The regression results show that the quality of literary translation = 0.165+0.124*Semantic retention + 0.356*Cultural appropriateness + 0.254*Linguistic fluency - 0.001*Content comprehensiveness + 0.256*Text accessibility. According to the results of the analysis, the impact of content comprehensiveness on the translation quality of literary works is not significant, and cultural appropriateness has the greatest impact on the translation quality of literary works.