We investigate how Korean-speaking preschool children develop clause-level constructional knowledge in expressing a transitive event (active transitives; suffixal passives) through Bayesian modelling. Adapting the Alishahi and Stevenson’s (2008) learning algorithm, we conducted a simulation by allowing our Bayesian model to learn constructional patterns as schematised input : pairings of morpho-syntactic and semantic-functional properties involving these constructions (with varying degrees of omission of sentential components). For this purpose, we devised artificial input based on characteristics of caregiver input manifested in CHILDES, and we measured posterior probabilities of each pattern per learning to estimate the degree of clustering for these constructions. Overall, we found dominance of several patterns and their inhibitory effects on the growth of the related patterns. Our findings suggest that constructional knowledge involving a transitive event proceeds from interactions between input properties and domain-general learning capacities, adding to cross-linguistic evidence for the effectiveness of Bayesian modelling on representing human learning.