In the context of a machine learning project, your team is tasked with preprocessing textual data by transforming words into numerical vectors. Which Databricks MLlib-supported technique is most suitable for this text-to-numerical conversion?