Structure mapping engine

Summary

SME is an implementation of the psychological theories of Gertner [1983]. It is an analogical matching algorithm that produces mappings between parts of source and target representations. As of 1990 there were over 40 projects used it [Falkenhainer et. al., 2005]. In a more recent review, French said Structure Mapping Theory is “unquestionably the most influential work to date of the modeling of analogy-making” [2002]. SME is useful because it ignores surface features and finds matches between potentially very different devices 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 go about it very differently. Structure Mapping Theory is based on the systematicity principle, which states that more connected knowledge is preferred over independent facts. Therefore, SME should ignore isolated source-target mappings unless they are part of a bigger structure. SME should map objects that are related to knowledge already mapped. Structure Mapping Theory also requires that mappings be one-to-one, which means that no part of the source description can map to more than one item in the target and no part of the target description can map to more than one part of the source. In addition, structure mapping theory requires that if a match maps S to T then the arguments of S and T must also be mapped. If both these conditions are met, the mapping is said to be structurally consistent.

---

Algorithm Details

The algorithm has several steps

Local Match Construction

fill in the other steps here

---

---

---

References

Papers by the Qualitative Reasoning Group at Northwestern University