Discovery of Analogical Knowledge for the Transfer of Workflow Tasks

Analogical knowledge considers functional properties of objects in contrast to literal similarity which compares the degree of featural overlap. A classical example from Gentner's structure mapping theory is "An electric battery is like a reservoir" [1]. Acquiring analogical knowledge in a computational approach is a challenging task. In this paper, we present a solution that combines learning with knowledge engineering. The proposed knowledge discovery approach uses word embeddings to learn analogy on workflow tasks. The resulting knowledge is integrated with an ontology for the purpose of workflow transfer across application domains. A case study is conducted on the two example domains 'passenger and baggage handling at the airport' and 'SAP warehouse management'. The experimental results on comparing the computational analogy with a golden standard from a knowledge engineering expert are quite promising and provide a proof-of-concept for the feasibility of the approach.