In AI and philosophy, case-based reasoning (CBR) is a strategy for tackling novel problems by learning from the approaches taken to similar issues in the past.

"As more and more AI is entering the world, more and more emotional intelligence must enter into leadership." - Amit Ray.

CBR can be traced back to the early 1980s work of Roger Schank and his students at Yale University. Schank's dynamic memory model was the foundation for the first CBR systems, CYRUS by Janet Kolodner and IPP by Michael Lebowitz.

In the 1980s, several schools of CBR and closely related subjects emerged, focusing on legal reasoning, memory-based reasoning (a method of reasoning from instances on massively parallel machines), and combining CBR with other reasoning approaches. In the 1990s, international interest in CBR developed, as indicated by the founding of an International Conference on Case-Based Reasoning in 1995, as well as workshops in Europe, Germany, the United Kingdom, Italy, and other countries.

Where is it used?

Case-based reasoning is used in everyday life by an auto repair which fixes an engine by recalling another car with similar symptoms. In addition, case-based reasoning is used by a lawyer who promotes a particular outcome in a trial based on legal precedents or by a judge who creates case law. 

Biomimicry

Similarly, an engineer who mimics the functional aspects of nature (biomimicry) treats nature as a database of problem-solving solutions. A common type of analogy solution generation is case-based reasoning. 

A help desk is a typical example of a case-based reasoning system. Users call the help desk with problems that need to be answered. The diagnostic assistant could use case-based thinking to help users determine what's wrong with their computers.

Case-based reasoning is an effective way of computational thinking and typical behaviour in ordinary human issue-solving. But, more radically, it is also that all reasoning is based on personally experienced past situations. This point of view is similar to prototype theory, which has received the most significant attention in cognitive science.

In general, the case-based reasoning method comprises the following steps: 

• Retrieve - It brings to mind an experience related to the current predicament. 
• Reuse - Reuse entails proposing a solution based on previous experience and changing it to meet the needs of the new circumstance. 
• Revise- It evaluates the solution's use in the new context.
• Retain- Once the solution has been successfully adapted to the specified problem, the resulting experience should be stored as a new case in memory.

Benefits

The value of CBR resides in the fact that it streamlines the learning process and results in better problem-solving. In case-based reasoning, the first step is to formulate the case's retrieval characteristics in light of the current objective. The second is to use the retrieval feature to look for similar issues in the memory case base.

Application

The application of CBR technology has resulted in the deployment of several successful systems, the first of which is Lockheed's CLAVIER, a method for laying down composite parts to be baked in an industrial convection oven. In addition, CBR has found widespread application in applications such as the Compaq SMART system and in the health sciences and structural safety management.

Recent work formalizes case-based inference as a probabilistic inference and develops CBR inside a statistical framework. As a result, case-based forecasts with a given level of confidence become possible. One way to distinguish CBR from induction from instances is that statistical inference seeks to discover what tends to make cases similar. In contrast, CBR seeks to encode what suffices to claim similarly.

Image source: Unsplash