Abstract
Described in this paper is a logic-modeling framework, based on fuzzy neural networks (FNNs) to identify causal relationships among variables. A case study is presented from the industrial construction domain to demonstrate the capability of the proposed system in finding plausible explanations of observed performance failures.