Case-based reasoning

The case-based Close (English case- based reasoning, French raisonnement par cas, Spanish razonamiento basado en casos ) is a machine learning method for problem solving by analogy. The central element in a CBR system is a so-called base case ( case database, samsung memory), in the solved problems are already stored as a case. One such case consists at least of a problem description and an associated problem solving. The goal is to use the solution of a similar and earlier already solved the problem of solving a given problem. This one imitates a human behavior by: Faced with a new problem, the person remembers often to a similar situation that he has experienced in the past, and tries, similar to master the task at hand.

Occasionally, one also speaks of remembrance based reasoning.

Method

The best-known model is the scientist Agnar Aamodt and Enric Plaza back that have the basic principle of case-based reasoning as a process with four phases, the CBR cycle known as described (Source: see below).

Application

Case- based reasoning has been found particularly in application systems for customer service, so-called help desk systems, proven, where you use it for diagnosis and treatment of customer problems, for example. More recently, it is amplified in (product) guidance systems a, for example, in e-commerce, as well as for the classification of texts.

Is considered advantageous that CBR can be applied to poorly structured and incomplete problems described. Unlike neighboring concepts (see below) is sufficient initially have a relatively small collection of reference cases, which increases by working with the CBR system gradually. Also in application domains, the precise effect relationships are not fully known, CBR is suitable.

As always, if one argues with analogies, it must be ensured that the system-generated solutions to the problem are adequate, so if for example, the conditions on which was based the historical solution, are still met, etc. ( obsolescence of knowledge ).

Classification

The Case- Based Reasoning is a branch of Artificial Intelligence and herein may be counted among the machine learning methods. The learning process is based on analogy, in contrast to learning by induction and deduction. Due to numerous applications in companies (see above) one is occupied not only in the (core) computer science but also in business computer science with CBR.