Artificial intelligence has come a long way from predicting simple weather temperatures using meteorological factors to generating poems and songs from a single line of text. Natural Language Processing as a domain of AI started by predicting sentiment analysis from statements. Such problems dealt with text inputs as complete sentences and a number as an output, which determined the positive or negative tone of the sentences, that is, a simple classification problem. Once we could determine or categorize text based on some classes, we could apply the same technique to other problems, such as statement comparisons that determine whether two statements are similar or different. There again, the input is a combination of text (two sentences) and the output is a number determining the class. Taking this a step further, a similar technique can be applied to extract intent and entities from text. Intent and entities are basic building blocks of chatbots and you will see them in detail in a later chapter.