Need help to validate this workflow approach

Hi I am performing NLP Text Classification Prediction model in KNIME.

This is my workflow

Description:

• Training + Test data contains 110 rows

• Deployment data contains 30 rows

• Initially I encountered without ‘concatenate’. Error is “ column is missing”

• To mitigate the issue, I used ‘Concatenate’ with Union as the option. Model prediction was successful wi th 94% accuracy.

Questions)

1. Is this the expected way to solve such text classification problem with new fresh deployment data?

2. Am I missing anything here?