In the most general sense, abstraction is the process of removing characteristics from something in order to reduce it to a set of essential characteristics. In computer science, abstraction is used to hide details of implementation in order to allow the focus to be on the interface rather than the implementation. For example, when a programmer uses a library function, they generally do not need to know how that function is implemented, only how to call it and what it does. This allows programmers to work at a higher level and makes code more reusable. Similarly, in object-oriented programming, abstraction is used to hide the implementation details of an object and instead provide a set of methods that can be used to manipulate the object. This allows objects to be treated as black boxes, which makes code more robust and easier to maintain.

Abstraction in artificial intelligence (AI)

Abstraction is a widely used concept in artificial intelligence to manage the use of different levels of detail in a representation language or the ability to switch between levels while preserving important characteristics. Its study has been mainly focused on problem-solving, theorem proving, knowledge representation (especially for spatial and temporal reasoning), and machine learning. In these contexts, abstraction is defined as a mapping between formalisms that reduces the computational complexity of the task being considered. By analyzing the concept of abstraction from an information quantity perspective, we can identify the differences and complementary roles of reformulation and abstraction in any representation change.
Abstraction is a critical mechanism that underlies human and artificial perception, knowledge representation, reasoning, and learning. It is a fundamental concept in various fields, including computer programming, natural and artificial vision, complex systems, artificial intelligence and machine learning, art, and cognitive sciences.