SHAP: How to get feature importance for categorical variable?

Question

I am trying to use the SHAP for an ANN model interpretation. I found that when I using

shap.summary_plot(shap_values[0], backgournd, plot_type=\'bar\')