Case-based reasoning

Case-based reasoning (CBR) is a concept for problem solving based on solutions for similar problems. The central module of a CBR system is the case memory that stored the previously solved problems in form of a problem description and a problem solution. One of the first comprehensive approaches for this method was described by Aamodt and Plaza in 1994 [16]. The process of case-based reasoning consists of four steps  

Cases can be represented in a variety of formalisms like frames, objects, predicates, semantic nets, and rules. Cases are usually indexed to allow fast and efficient retrieval. Several guidelines on indexing have been proposed by CBR researchers [17,18]. Both manual and automated methods have been used to select indices. Choosing indices manually involves deciding a case s purpose with respect to the aims of the reasoner and deciding under what circumstances the case will be useful. Indices can be automatically constructed on several paradigms  

Indices may also be created by using inductive learning methods — for instance, artificial neural networks — for identifying predictive features that are then used as indices. However, despite the success of many automated methods, Janet Kolodner believes that people tend to do better at choosing indices than algorithms, and therefore for practical applications indices should be chosen manually [19]. 

The case base is organized in a structure that allows efficient search and retrieval. Several case-memory models have been proposed. The two most widely used methods are the dynamic memory model of Schank and Kolodner [20,21] and the category-exemplar model of Porter and Bareiss [22]. 

If the ES does not have sufficient information to be able to answer a user s query using only information from its rule base and repository of facts, it may refer to the contents of its case library. 

The aim of case-based reasoning is to provide advice based on a set of known examples that are judged to be relevant to the user s query. Files within the library contain data about past cases relevant to the area of expertise, how they were tackled, what the results of this approach were, and whether the action taken was appropriate and successful. Each case is tagged with a set of attributes that describe the case, so that when the library is searched for relevant material, it can quickly be identified through some form of similarity metric. 

If a case is found that matches the user s query closely, this is used to provide appropriate advice. If, on the other hand, there is no case that matches the problem presented by the user sufficiently closely, the system can try to modify a case that is present in the library to bring it into line with the user s query in what is known as structural adaptation, or it may be able to create a new solution using as a starting point a similar case from the past (derivational adaptation). 

If neither approach works, but the problem is eventually solved through discussion with the user, the system can add the current case and its solution to the library, thus expanding the system s understanding of relevant cases. 

---

Case-Based Reasoning. Case-Based Reasoning (CBR) systems base their solutions on previously solved problems (cases) which are stored in a case-base [Watson Marir, 1994]. When a new problem is presented to a CBR system a similar case(s) is/are retrieved from the case-base. Depending on the differences between the retrieved and the presented problem the retrieved solution may have to be more or less adapted to obtain a solution to the new problem. The solved problem may be retained in the case base if deemed useful. 

Case-based reasoning. The main advantage of CBR systems for NDT data interpretation is that they can cope with data coming from inspection of varying constructions under varying conditions with various system settings due to their ability to learn from the data classified by the operator. In such situations no reliahle statistical classifier can be designed, and the rule-hased classifiers would be either very inefficient or unpractically complex. 

Case-based reasoning CA. See Cyanuric acid Case-hardened steels Case hardening... 

Eor a number of cognitive or interpretive tasks, there are alternatives to mainstream knowledge-based systems that may be more appropriate, especially if adaptive behavior and learning capabihty are important to system performance. Two approaches that embody these characteristics are neural networks (nets) and case-based reasoning. 

K. Sycara, "Using Case-Based Reasoning for Plan Adaptation and Repair," in Proceedings of the DARPA Workshop on Case-Based Reasoning, Morgan... 

D. Navinchandra, "Case-Based Reasoning in CYCLOPS, A Design Problem Solver," in Ref. 82. 

P. Koton, "SMARTPLAN A Case-Based Resource Allocation and Scheduling System," in Proceedings of a Workshop on Case-Based Reasoning, Morgan Kaufman, San Mateo, Calif., 1989. 

W. Mark, "Case-Based Reasoning for Autoclave Management," in Ref. 85. 

Case-based reasoning is very much dependent on the structure and content of its cases and their representation because case retrieval involves identifying those features in the problem that best match those in the case base. The dynamic addition of new cases means that CBR is intrinsically a learning methodology such that the performance of an expert system based on this approach will improve with time [9]. Systems may be developed with conventional computer languages or shells [7]. 

Goodall A. Preface. In Althoff KD, A review of industrial case-based reasoning tools. Oxford AI Intelligence, 1995. 

Aamodt A, Plaza E. Case-based reasoning—foundational issues, methodological variations and system approaches. AI Commun 1994 7 39-59. 

Watson I. Applying case-based reasoning techniques for enterprising systems. San Francisco Morgan Kaufmann, 1997. 

Rowe RC, Craw S, Wiratunga N. Case-based reasoning—a new approach to tablet formulation. Pharm Tech Eur 1999 ll(2) 36-40. 

And case-based reasoning, in which the decision on how to act in one situation is guided by what action was successful when a similar situation was encountered in the past, is also of potential value ... 

Keywords inherent safety, process plant design, safety analysis, case-based reasoning, genetic algorithm... 

A new approach for computerized Inherent Safety Index is also presented. The index is used for the synthesis of inherently safer processes by using the index as a fitness function in the optimization of the process structure by an algorithm that is based on the combination of an genetic algorithm and case-based reasoning. Two case studies on the synthesis of inherently safer processes are given in the end. 

When problem solving is based on experience which is difficult to define as explicit rules, it is possible to apply case-based reasoning (CBR). CBR uses directly solutions of old problems to solve new problems. The functional steps in CBR are (Gonzalez and Dankel, 1993) ... 

Case-based reasoning has earlier been used for instance for equipment design. Koiranen and Hurme (1997) have used case-based reasoning for fluid mixer design and for the selection of shell-and-tube heat exchangers. They have included an estimation of design quality for the case retrieval beside technical factors. 

The fitness function was based on the inherent safety index, which was simplified It was noticed that there are only minor differences in the safety properties of the compounds in the process. Therefore most subindices are the same for all configurations. The equipment type used in all the configurations is the same (i.e. distillation). Therefore the subindex of equipment safety is constant too. Also the safety of process structures is quite the same since the distillation systems used are rather similar in configuration. Therefore the subindex for process structure was not evaluated and case-based reasoning was not needed. 

Case-based reasoning on the safe process structure subindex... 

The chemical and most process factors affecting the index are quite straightforward to estimate. More problematic are the equipment safety and the safety of process structure. The equipment safety subindex was developed based on evaluation of accident statistics and layout information. The evaluation of the safe process structure subindex is based on case-based reasoning, which requires experience based information on accident cases and on the operation characteristics of different process configurations. 

In design it is typical that the same mistakes are done again since the use of available information is not organized. The use of case-based reasoning enhances the reuse of available experience, which reduces the possibility that the same errors are done more than once. In this work CBR was used for the evaluation of the inherent safety of process structure. The casebase was collected from design standards, accident documents and good engineering practice. 

Heikkila, A-M., Koiranen, T. Hurme, M. 1998. Application of Case-Based Reasoning to Safety Evaluation of Process Configuration. Rugby Institution of Chemical Engineers. IChemE Symposium Series No 144, pp. 461-473. 

Hurme, M. Heikkila, A-M. 1998. Synthesis of Inherently Safe Chemical Processes by Using Genetic Optimization and Case-Based Reasoning. In Koikkalainen, P. Puuronen, S. (Eds.). Human and Artificial Information Processing. Finnish Artificial Intelligence Society. Espoo. Pp. 134—143. 

Koiranen, T. Hurme, M. 1997. Case-Based Reasoning Application in Process Equipment Selection and Design. Sixth Scandinavian Conference on Artificial Intelligence, G. Grahne (Ed.), Amsterdam IOS Press. Pp. 273-274. 

Kolodner, J. 1993. Case-Based Reasoning. San Mateo Morgan Kaufman Publishers. 

Most AI methods used in science lie within one of three areas evolutionary methods, neural networks and related methods, and knowledge-based systems. Additional methods, such as automated reasoning, hybrid systems, fuzzy logic, and case-based reasoning, are also of scientific interest, but this review will focus on the methods that seem to offer the greatest near-term potential in science. 

---