Importance of vectorization in deep learning

This is more of a discussion with you from my end. As we all know, computers can't understand NL directly, so we need to convert our NL output into numerical format. We have various word embedding techniques, as well as some basic statistical techniques such as indexing, tf-idf, one-hot encoding, and so on. By using all these techniques, or some of these techniques, you can convert your text input into numerical format. Which techniques you choose totally depends on the NLP applications. So, there are two major points behind why we convert NL input to numerical format. It is basically done because the computer can only understand numerical data, so we have to convert text data to numerical data and computers are very good at performing computation on given numerical data. These are two major points that come to my mind when we...