Kedar-Cabelli, Smadar T. Analogy -- from a unified perspective. Technical Report MLTR -3, Rugters University Laboratory for Computer Science Research, December 1985.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....features from arbitrary figures is beyond the scope of this paper. This paper also can be thought of as a study of specifically visual analogical reasoning. Most work to date on analogical and case based reasoning has assumed a fundamentally sentential representation of the models in question (Kedar Cabelli, 1985). However, the special characteristics of visual representations in human and computer problem solving is a growing area of research (Glasgow et al. 1995) Future questions along those lines that bear investigation include, ffl What is the relationship between sentential analogy and visual ....

Kedar-Cabelli, Smadar T. Analogy -- from a unified perspective. Technical Report MLTR -3, Rugters University Laboratory for Computer Science Research, December 1985.

....is a notoriously difficult term to define. This confusion arises mainly because analogy means different things to different researchers [Russell, 89: pages 112132 ] Etymologically, the word analogy comes from the Greek word analogia meaning mathematical proportion (such as 1:2 : 2:4) Kedar Cabelli, 88: page 65] but now, the word has a broader meaning. A psychology based definition of analogical reasoning can be found in [Reber, 85] Chapter 2 : Analogical Reasoning 13 analogical reasoning, reasoning whereby decisions about objects, events or concepts depend upon perceived similarities in ....

Kedar-Cabelli, S. "Analogy - From a unified perspective." Analogical Reasoning : Perspectives of Artificial Intelligence, Cognitive Science and Philosophy. Ed: D. H. Helman. Dordrecht, Netherlands: Reidel, 1988. : 65-103.

....critics: Yet ANALOGY has its serious limitations. Using the domain of geometric figures is the most serious limitation of the program. Because of the many hidden assumptions carried along with such a domain, AI work in analogical reasoning has been misdirected in a number of critical ways. (Kedar Cabelli, 1988) Of course, as ANALOGY doesn t learn, I, too, think that ANALOGY has its limitations. However, I disagree with Kedar Cabelli as to what those problems are. She states that the domain of geometric figures has misdirected the entire study of analogymaking because of many hidden assumptions. By ....

Kedar-Cabelli, S. (1988). Analogy---from a unified perspective. In D.H. Helman (Ed.), Analogical Reasoning. Kluwer.

....many researchers. For example, the use of analogy in problem solving and learning was proposed by, among others, Carbonell (Carbonell and Michalski, 1983a; Carbonell and Michalski, 1983b; Carbonell, 1986) Winston (Winston, 1980) Falkenhainer et al. Falkenhainer and Forbus, 1989) Kedar Cabelli (Kedar Cabelli, 1988), Burstein (Burstein, 1986) and Kodratoff (Kodratoff, 1988) 1 One of the above (Winston, 1984) demonstrates the importance of past experience as being one element in a hypothetical intelligent machine. Winston stresses the importance of past experience in learning, where people use their ....

....(Rumelhart, 1981) and Gentner (Gentner, 1983) provide examples of such theories. Gentner (Gentner, 1982) uses well defined relations in a 1 A detailed survey to the computational approaches to analogical reasoning is found in, among others, Ellman, 1989) Hall, 1989) Kodratoff, 1988) and (Kedar Cabelli, 1988) Chapter 8. Application: Analogical reasoning and EIR 178 rich domain to understand and explain relations that exist in a poor domain. Most of these analogical reasoning models have been developed to form a kind of partial pattern matching. The mechanisms of these models are based on the ....

[Article contains additional citation context not shown here]

Kedar-Cabelli, S. (1988). Analogy-from a unified perspective. In Helman, D., editor, Analogical reasoning-Perspectives of Artificial intelligence, cognitive science and philosophy, volume 197, pages 65--103. Kluwer academic publishers.

....similarity for reuse matching and retrieval is that it could provide some sort of analogical reasoning for component matching. However, in such systems only certain kinds of similarity would be relevant to the desired inferences while others could be totally irrelevant or even even misleading [Kedar Cabelli 1988]. Work in NATURE will first determine candidate dimensions along which similarity could be exploited in requirements engineering (e.g functional similarity, structural similarity, complexity, similarity on reliability) then evaluate these candidates. Empirical investigations into the influence of ....

Kedar-Cabelli S., Analogy - from a unified perspective, Analogical Reasoning, Kluwer Academic Publishers, 1988

....his own productivity, and also to improve his own learning. A system of this kind should also support reuse by analogy. It is in this context we study analogical reasoning and analogical problem solving. It is agreed that the analogical problem solving process can be divided into four phases [12, 20]: A retrieval phase, a matching phase, an adaption phase, and a learning phase. The last three of these phases have been investigated to some extent (for instance [5, 9, 15] whereas very little research [7] has been concentrated on base filtering, which is the term coined for the first step. For ....

S. Kedar-Cabelli. Analogy---from a unified perspective. In D.H. Helman, editor, Analogical Reasoning, pages 65--103. Kluwer Academic Publishers, 1988.

....filters; as our number of possibilities decreases, we could look for discriminating measures which could further reduce the set, just as one incrementally adds indices to a document retrieval in order to filter out the unwanted documents. At what level of abstraction do we retrieve Kedar Cabelli [15] has noted that we cannot retrieve based on surface features alone, but must consider shared abstractions. How can we look at shared abstractions efficiently It should be noted that psychological evidence shows that, while people do use abstractions for retrieval, the retrieval process is largely ....

Kedar-Cabelli, S.T., "Analogy from a Unified Perspective," Report No. ML-TR-3, Rutgers University, New Brunswick, NJ, December 1985.

.... 1986) methods to generate and use plausible explanations in an incomplete knowledge base (Schank et al. 1986) Lenat and Guha, 1989) knowledge intensive case based learning case based reasoning (Hammond, 1989) Koton, 1989) Porter et al. 1990) and analogical reasoning and learning methods (Kedar Cabelli, 1988), Kodratoff, 1990) Apprenticeship learning. The notion of learning apprentice systems was introduced in (Mitchell et al. 1985) as interactive knowledge based consultants that directly assimilate new knowledge by observing and analyzing the problem solving steps contributed by their users ....

....1986) also show the frequent use of past experience in solving new and different problems. Case based reasoning and analogy are 15 sometimes used as synonyms (e.g. by Carbonell) viewing CBR intra domain analogy. However, as will be discussed later, the main body of analogical research (Kedar Cabelli, 1988), Hall, 1989) have a different focus, namely analogies across domains (Burstein, 1989) 4.2. Main types of CBR methods. Case based reasoning is a broad term, covering many particular type of methods. Below is a list of different CBR methods, distinguished by their different solutions to core CBR ....

[Article contains additional citation context not shown here]

Kedar-Cabelli S. (1988). Analogy - from a unified perspective. In: D.H. Helman (ed.), Analogical reasoning. Kluwer Academic. pp 65-103.

....Finally, some experimental results on synthetic data are presented in section six. 2. Analogical reasoning Analogical reasoning in artificial intelligence, as presented by most researchers on the topic, can be split into several phases; for instance retrieval, mapping, transfer, and learning [11], 8] 15] Retrieval is the process of searching through a collection of situations to find a potential analogue (base) to the situation at hand (target) Mapping is to establish the correspondencies between objects and relations in the base and the target. Transfer is to establish new ....

Smadar Kedar-Cabelli. Analogy---from a unified perspective. In D.H. Helman, editor, Analogical Reasoning, pages 65--103. Kluwer Academic Publishers, 1988.

....the frequent use of past experience in solving new and different problems. Casebased reasoning and analogy are sometimes used as synonyms (e.g. in [16] Case based reasoning can be considered a form of intra domain analogy. However, as will be discussed later, the main body of analogical research [14,23,30] has a different focus, namely analogies across domains. In CBR terminology, a case usually denotes a problem situation. A previously experienced situation, which has been captured and learned in such way that it can be reused in the solving of future problems, is referred to as a past case, ....

....methods focus on indexing and matching strategies for single domain cases. Research on analogy reasoning is therefore a subfield concerned with mechanisms for identification and Aamodt and Plaza: Case Based Reasoning AICOM Vol. 7 Nr. 1 March 1994 6 utilization of cross domain analogies [23, 30]. The major focus of study has been on the reuse of a past case, what is called the mapping problem: Finding a way to transfer, or map, the solution of an identified analogue (called source or base) to the present problem (called target) Throughout the paper we will continue to use the term ....

Kedar-Cabelli, S. (1988): Analogy - from a unified perspective. In: D.H. Helman (ed.), Analogical reasoning. Kluwer Academic, 1988. pp 65-103.

....Work by Gentner, Holyoak and Thagard, and work on case based reasoning will be reviewed later in this section. There has also been attempts to represent analogical reasoning in formal languages [22, 18, 19] but much remains in order to make these languages complete. According to Kedar Cabelli [26] computational models for analogical problem solving can be described in 4 steps: 1. Retrieval: From a target case, find a base case in the knowledge base that resembles the target and has a reusable solution. 2. Elaboration: Derive attributes, relations and causal chains that involve the base ....

....should prefer mapping causal networks of relations over independent (unrelated) relations. By preferring causal networks, we constrain the number of possible mappings between base and target for any analogy. Although recognized as important, Gentner s work has received a fair amount of criticism [21, 26]. We will not elaborate this further, but only mention that one of the main drawbacks is that the structure mapping theory assumes that all causal relations describing the base are candidates for mapping. It is certainly difficult to distinguish the relevant causal relations from the irrelevant. ....

[Article contains additional citation context not shown here]

Smadar Kedar-Cabelli. Analogy---from a unified perspective. In D.H. Helman, editor, Analogical Reasoning, pages 65--103. Kluwer Academic Publishers, 1988.

No context found.

Kedar-Cabelli, S. (1988). Analogy---from a unified perspective. In D.H. Helman (Ed.), Analogical Reasoning. Kluwer.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC