Case-Based Reasoning

 Case-Based Reasoning (CBR) is motivated by observing how humans reason. When faced with a problem situation, humans often remember how they were able to solve a similar situation, rather than to re-think the entire problem. Case-Based Reasoning is decision support which uses earlier experiences, or cases, as the basis to make decisions. It can be defined as ``to solve a problem, remember a similar problem you have solved in the past and adopt the old solution to solve the new problem'' [6]. A more formal definition is given in [12]:

Case Based Reasoning is a problem solving paradigm which utilizes the specific knowledge of previously experiences, concrete problem situations (cases). A new problem is solved by finding a similar past case, and reusing its solution in the new problem situation. CBR is a cyclic and integrated process of solving a problem and learning from this experience.

An overview of the technology is given in [22].

A case is an abstraction of an event, limited in time and space. It can either be a previous encountered and solved problem, or a typical way of solving a problem. The differences in notation will be evident from the different methods of CBR, outlined in the next subsection.

Althof, Auriol et al. [6] see a case as consisting of:

• A description
• An associated solution
• The justification for the solution

Where the justification is an explicit representation of the problem solving process. It can refer to other cases, textual explanation or to general rules. Cases can be described in a variety of languages, and be represented as flat data tables or with a structure such as a tree. The cases can be positive or negative experiences.