This approach is useful when you have categorical data. The intuition behind this approach is that missing values may correlate with other variables, and removing them will result in a loss of information that can affect the model significantly.
For example, if we have a binary variable with two possible values, -1 and 1, we can add another value (0) to indicate a missing value. You can use the following code to replace the null values of the Cabin feature with U0:

# replacing the missing value in cabin variable "U0"
df_titanic_data['Cabin'][df_titanic_data.Cabin.isnull()] = 'U0'