Chapter 5 Natural interaction: Can computers understand humans and talk to them? 
AI is now used to make machines intelligent and interact and communicate like humans. For example, we can mention the case of Siri which is a smart virtual assistant developed by Apple Inc (see several examples in Figure 27). It uses voice queries and a natural-language user interface to answer questions, make recommendations, and perform actions by delegating requests to a set of internet services.
To conduct natural language interaction and communicate with humans, a machine conducts the following steps: (1) take the human input, such as speech; (2) analyze the input (speech) to understand what a person is trying to say; (3) do the reasoning process to provide an accurate answer or perform a task; and, (4) perform the action, by answering the user back or executing a specific task. To deliver an intelligent, humanlike experience, a machine should use a number of Natural Language Processing (NLP) principles and technologies. For instance, Figure 28 shows a communication scenario between a human and a virtual assistant to play a specific music.
NLP is defined as the application of computational techniques to the analysis and synthesis of natural language and speech. In other words: the use of different techniques from computer science (algorithms) to understand and manipulate human language and speech. NLP has the following two main components:
Natural Language Understanding (NLU): It revolves around machine reading comprehension. This is an AI-hard problem. An NLU system needs the following components: (1) Lexicon, Parser, and Grammar rules; and, (2) Semantic theory to guide comprehension.
Natural Language Generation (NLG) is concerned with generating natural language. It uses a machine representation system like a knowledge base or a logical form. You can think of it as a translator between data and natural language representation; this is the opposite of NLU. This involves three tasks: (1) Text Planning- To extract relevant content from the knowledge base; (2) Sentence Planning- To choose appropriate words, form meaningful phrases, and set sentence tone; and, (3) Text Realization- To map the sentence plan into sentence structure.