Tokenization is a essential method in natural language processing (NLP), acting as the initial point for preparing text data. Essentially, it involves breaking down a larger block of text – like a paragraph – into smaller units called “segments". These tokens can be copyright, special characters, or even sub-lexical segments. This conversion