In artificial intelligence and cognitive science, the structure mapping engine (SME) is an implementation in software of an algorithm for analogical matching based on the psychological theory of Dedre Gentner.
The basis of Gentner's structure-mapping idea is that an analogy is a mapping of knowledge from one domain (the base) into another (the target).
[1] The theory is useful because it ignores surface features and finds matches between potentially very different things if they have the same representational structure.
For example, SME could determine that a pen is like a sponge because both are involved in dispensing liquid, even though they do this very differently.
Structure mapping theory is based on the systematicity principle, which states that connected knowledge is preferred over independent facts.
Entities represent the objects or concepts in a description — such as an input gear or a switch.
t) The predicate's functor is “behavior-set,” its type is “relation,” and its n-ary and commutative parameters are both set to true.
[2] The first step of the algorithm is to create a set of match hypotheses between source and target dgroups.
These rules are not the place where domain-dependent information is added, but rather where the analogy process is tweaked, depending on the type of cognitive function the user is trying to emulate.
This limitation makes the processing more efficient by constraining the number of match hypotheses that are generated.
In order to illustrate how the match rules produce match hypotheses consider these two predicates: transmit torque inputgear secondgear (p1) transmit signal switch div10 (p2) Here we use true analogy for the type of reasoning.
The intern rules then produce three more match hypotheses: torque to signal, inputgear to switch, and secondgear to div10.
If the arguments were functions or attributes instead of entities, the predicates would be expressed as: transmit torque (inputgear gear) (secondgear gear) (p3) transmit signal (switch circuit) (div10 circuit) (p4) These additional predicates make inputgear, secondgear, switch, and div10 functions or attributes depending on the value defined in the language input file.
The reason SME does not match attributes is because it is trying to create connected knowledge based on relationships and thus satisfy the systematicity principle.
For example, if both a clock and a car have inputgear attributes, SME will not mark them as similar.
Once the match hypotheses are generated, SME needs to compute an evaluation score for each hypothesis.
Multiple amounts of evidence are correlated using Dempster's rule [Shafer, 1978] resulting in positive and negative belief values between 0 and 1.
Rule 5 provides trickle-down evidence in order to strengthen matches that are involved in higher-order relations.
The rest of the SME algorithm is involved in creating maximally consistent sets of match hypotheses.
SME then uses this information to merge match hypotheses — using a greedy algorithm and the structural evaluation score.
Chalmers, French, and Hofstadter [1992] criticize SME for its reliance on manually constructed LISP representations as input.
They argue that too much human creativity is required to construct these representations; the intelligence comes from the design of the input, not from SME.
Turney [2008] presents an algorithm that does not require LISP input, yet follows the principles of Structure Mapping Theory.
Turney [2008] state that their work, too, is not immune to the criticism of Chalmers, French, and Hofstadter [1992].
This leads to the prediction that analogy making proceeds not by mapping correspondences from candidate sources to target, as predicted by the structure mapping theory of analogy, but by weeding out non-correspondences, thereby whittling away at potentiality."