Case-Based Reasoning for Construction Hazard Identification: Case Representation and Retrieval

• Published in: Journal of construction engineering and management Vol. 135; no. 11; pp. 1181 - 1189
• Main Authors: Goh, Y. M, Chua, D. K. H
• Format: Journal Article
• Language: English
• Published: Reston, VA American Society of Civil Engineers 01-11-2009
• Subjects: Applied sciences; Buildings. Public works; Exact sciences and technology; Project management. Process of design; TECHNICAL PAPERS; Project management; Knowledge base; Case based reasoning; Modeling; Information technology; Construction industry; Case study; Accident; Information technology (IT); Safety; Knowledge-based systems; Risk management; Construction management; Accidents
• ISSN: 0733-9364; 1943-7862
• DOI: 10.1061/(ASCE)CO.1943-7862.0000093

This paper proposes a case-based reasoning (CBR) approach to construction hazard identification that facilitates systematic feedback of past knowledge in the form of incident cases and hazard identification. This paper focuses on two of the key components of the CBR approach: (1) a detailed knowledge representation scheme, developed based on the modified loss causation model, to codify incident cases and past hazard identification and (2) an intelligent retrieval mechanism that can automatically retrieve relevant past cases. The detailed knowledge representation scheme presented herein is designed to model both incident cases and hazard identification so that both types of knowledge repository can be retrieved simultaneously and adapted for use. The scheme also includes a linguistic structure used to facilitate indexing of cases. The retrieval mechanism is based on the concept of similarity scoring. In this paper, a novel scoring technique based on semantic networks is presented. A case study is presented to demonstrate and validate the proposed approach.