Home » Machine Learning/Artificial Intelligence

Components of communicating agents | Artificial Intelligence

In this tutorial, we will learn about the components of communicating agent in artificial intelligence. By Monika Sharma Last updated : April 16, 2023

As discussed earlier, communicating consists of the speaker and the hearer. Because for communication to take place, the agent must be able to perform both these tasks. Both these components can be further explained as follows on the basis of their functioning:

Speaker

1. Intention:
   Before speaking anything, we know the intention of what we want to convey to the other person. The same thing is implemented in the communicating systems. This makes communication valid and relevant from the side of the communicating system.
2. Generation:
   After knowing the intention of what is to be conveyed, the system must gather words so that the information can be reached to the user in his very own communicating language. So, the generation of relevant words is done by the system after the intention process.
3. Synthesis:
   Once the agent has all the relevant words, yet they have to be uttered in a way that they have some meaning. So, after the generation of words, the formation of meaningful sentences takes places and finally, the agent speaks them out to the user.

Hearer

1. Perception:
   In the perception phase, the communicating system perceives what the user has spoken to it. This is a sort of an audio input signal which the agent receives from the user and then this signal is sent for the further processing by the system.
2. Analysis:
   After getting the audio input from the user which is a sequence of sentences and phrases, the system tries to analyze them by extracting the meaningful terms out of the sentences by removing the articles, connectors and other words which are there only for the sake of sentence formation.
3. Disambiguation:
   This is the most important thing that a communicating system carries out. After the analyzing process, the agent must understand the meaning of the sentences that the user have spoken. So, this understanding phase in which the system tries to derive the meaning of the sentences by removing various ambiguities and errors is known as disambiguation. This is done by understanding the Syntax, Semantics, and Pragmatics of the sentences.
4. Incorporation:
   In incorporation, the system figures out whether the understanding that it has derived out of the audio signal is correct or not. Whether it is meaningful, whether the system should consider it or ask the user for further input for resolving any sort of ambiguity.

---