“…The risk assessment methods that capture the interdependency between the risk events usually employ analytical network process (ANP) (Boateng et al , 2015), network theory (Fang et al , 2012), Bayesian belief network (Cárdenas et al , 2013; Yazdi and Kabir, 2017), simulation (Ou-Yang and Chen, 2017) and artificial neural networks (Barakchi et al , 2017; Kabir et al , 2016; Lee and Kim, 2017). Different other popular multi-criteria decision methods (Li et al , 2013) are also available for specific project to manage the complex risk networking such as FMEA (Lee and Kim, 2017), FTA (Khakzad et al , 2013), ETA (Ferdous et al , 2011), VIKOR (Kahraman et al , 2015), fuzzy set theory (Kimiagari and Keivanpour, 2018), AHP (Al-Harbi, 2001) and TOPSIS (Zegordi et al , 2012) etc., but these methods are not capable of capturing interdependencies between events. Consequently, the above mentioned artificial intelligence methods except ANP do not consider a pair-wise comparison between the criteria and sub-criteria of different events.…”