CA-FCM: Towards a formal representation of expert's causal judgements over construction project changes. (October 2018)

• Record Type: Journal Article
• Authors: Shahinmoghaddam, Mehrzad; Nazari, Ahad; Zandieh, Mostafa
• Is Part Of: Advanced engineering informatics. Volume 38(2018)
• Page Start: 620
• Page End: 638
• Publication Date: 2018-10
• Subjects: Fuzzy Cognitive Mapping (FCM) -- Contextual knowledge modelling -- Knowledge-based decision support system -- Semantic web technologies -- Construction project changes -- Causal knowledge representation; Computer-aided engineering -- Periodicals; Engineering -- Data processing -- Periodicals; 620.00285
• DOI: 10.1016/j.aei.2018.09.006
• Languages: English
• ISSNs: 1474-0346

Abstract

Highlights: CA-FCM is suggested for dynamic causal modeling of construction project changes. The knowledge-driven nature of the approach addresses data-scarcity issues at hand. A hybrid inference mechanism is embedded within the framework. CA-FCM's context-aware processes cope with context-sensitivities of causal judgments. Resulting shared sense-making of concepts facilitates manual construction of FCMs. A partial implementation of the proposed framework is presented. Abstract: Aimed at improving the proactive benefits of Fuzzy Cognitive Mapping (FCM) for predicting construction project changes, this paper presents CA-FCM: a Context-aware Fuzzy Cognitive Mapping approach. CA-FCM's main functionality is to imitate the intuitive causal judgements of project experts over change causation in different contextual settings. Invoking the logical inference capabilities of semantic web tools, a hybrid inference mechanism is embedded within the proposed framework which enables establishing contextual connections between prospective causal factors through a semi-automated process of generating relevant causal statements. Hence, CA-FCM can assist decision-makers with (1) a shared sense-making of the domain concepts which would significantly facilitate the manual construction of FCM scenarios, (2) providing contextualized recommendations of causal information required for developing FCM scenarios, (3) dynamic modelling of causal inferences, imitating expert reasoning on change causation and propagation. Towards providing a detailed delineation of CA-FCM's effectiveness on providing assistance in planning for project changes, a partial implementation of the proposed framework was conducted within a real case scenario.