Home » Machine Learning/Artificial Intelligence

Knowledge-Based Agent Levels in Artificial Intelligence

In this tutorial, we will learn about the knowledge-based agent. We will study the knowledge base, how it resides inside an agent and the different levels of the knowledge-based agent? By Monika Sharma Last updated : April 15, 2023

What are Knowledge-Based Agents?

The Knowledge-based agents are those agents which contains all the information the agent has. This information can either be the data that is embedded into the agent in prior while its manufacturing or it can be the knowledge that the agent had gained through the experiences. In the knowledge-based agent, all the decision are taken in accordance with the data present in its knowledge base.

Levels of Knowledge-based Agent

The knowledge-based agent can be described using three levels. These are:

1. Knowledge level
2. Logical level
3. Implementation level

1. Knowledge level

This is the basic and the most abstract level in a knowledge-based agent. This level describes the agent by what it knows, i.e. through its knowledge base. In this level, the information that the agent has, its goals and the utility are defined.

Example

A house cleaning agent knows where the mop is and where the floor is and where the cleaning material like the soap and detergents are kept. All this is defined in the knowledge level.

2. Logical level

In the logical level, the raw and discrete information which is present in the knowledge level is encoded into sentences. In simple words, the agent in this level derives the logic out of the knowledge base according to the problem. It selects the required blocks of information and combines them meaningfully so as a decision could be made.

Example

The house cleaning agent on getting all the information then checks and finds the place which is dirty and should be cleaned.

3. Implementation level

This is the final layer of the knowledge-based agent. In the implementation level, the logic which the agent has derived in the logical level is brought to implementation. This layer deals with the physical representation of the sentences.

Example

The cleaning agent after finding the place to clean finally takes action and cleans the dirt on the floor and as a result, the agent gets clean floor which is the goal state that the agent is designed to achieve.

---