Using text embedding models and vector databases as text classifiers with the example of medical data