bert-based-Chinese

his repository includes the data, code, and step-by-step instructions for reproducing the findings presented in our IJGIS manuscript, including Tables 1、3、4 and Figures 1–9. All files are anonymized and provided to support reproducibility.Section Figure/Table Script(s) Input Output3.1 Table 3 Code/bert.py data/wh_data_cleaned.csv Model accuracy assessment3.1 Table 4 Code/bert.py data/wh_data_cleaned.csv Examples of emotional structures3.1 Figure 4 Code/graph.py Result/bert/wh/128 Different types of emotions3.2 Figure 5 Code/mode_heatmap.py sentiment_density.csv Distribution of kernel density 3.3 Figures 7-8 Code/tree_v2.py Result/bert/wh/128/clustering_results.pkl Emotional grid management 4.2 Table 5-6 Code/tree_v2.py Result/bert/wh/128/clustering_results.pkl indexSection Figure/Table Source/Production Methods Output2.2 Table 1 data/wh_data_cleaned.csv Examples of data 2.2 Figure 1 ArcGIS pro Heatmap of the social media data 2.2 Figure 2 Customize images based on power point The proposed framework2.2 Figure 3 Customize images based on power point Fixed Grids vs. Adaptive Grids3.2 Figure 6 Data/emotion_prediction_wh.csv(Customize images based on EXCEL) Percentages of different emotional patterns4.1 Figure 9 Data/emotion_prediction_wh.csv(Customize images based on EXCEL) Histogram of comparisons among different types of emotions