Conversational Process-Oriented Case-Based Reasoning

摘要

Current approaches for retrieval and adaptation in process-oriented case-based reasoning (POCBR) assume a fully elaborated query given by the user. However, users may only have a vague idea of the workflow they desire or they lack the required domain knowledge. Conversational case-based reasoning (CCBR) particularly addresses this problem by proposing methods which incrementally elicit the relevant features of the target problem in an interactive dialog. However,-.no CCBR approaches exist that are capable of automatically creating;.questions from the case descriptions that go beyond attribute-value representations. In particular, no approaches exist that are applicable to workflow cases in graph representation. This paper closes this gap and presents a conversational POCBR approach (C-POCBR) in which questions related to structural properties of the workflow cases are generated automatically. An evaluation in the domain of cooking workflows reveals that C-POCBR can reduce the communication effort for users during retrieval.