Analogical Estimation

Analogy lets us make inferences from a better known example on to a less known one. Analogical estimation is a specific kind of analogical inference, namely, inferring the quantitative value of an unknown based on a known value from a similar example. We present Analogical Estimator, a computational model of analogical estimation. A key claim of this work is that we don’t need a separate mechanism to capture the role of quantitative values in similarity. We need better representations, specifically, symbolic representations of quantity that capture the essential qualitative distinctions