Focusing on the issue of national cultural connotation and narrative strategy in Chinese modern novels in the early 20th century, this paper introduces digital humanities and text computing methods, and constructs an analysis framework composed of national cultural implication representation, narrative strategy structural identification, multi-dimensional feature correlation analysis and parameter adaptive adjustment. Based on the corpus of representative modern novels from 1918 to 1932, this paper jointly identified the cultural dimensions and narrative bearing modes such as national consciousness, ethical order, subject awakening, custom memory through text cleaning, hierarchical annotation, feature extraction and model training. The results show that the Accuracy of the proposed framework reaches 91.4%, the F1 value reaches 89.8%, and the RMSE is reduced to 0.146 under the condition of a unified corpus, which is better than the fixed threshold model, BiLSTM model and Text-GNN model. Ablation experiments and robustness tests further show that multidimensional association analysis and parameter adaptation mechanism play an important role in stable interpretation in complex literary contexts. The results show that computer-aided analysis can provide comparable and retractable structural evidence for the study of national culture of modern Chinese novels in the early 20th century, and also provide a new way for the integration of modern literature research and digital humanities methods.
