Clique structure and node-weighted centrality measures to predict distribution centre location in the supply chain management

摘要

Much importance is attached to the weights on the edges in a network, but in actual fact what makes up a network is both the nodes and the edges linking up the network. It is therefore pertinent to investigate the effects and importance of the weights attributed unto the nodes in a network as well as the weights on the links of such networks as they both play important roles in determining the prominence or popularity of actors within any particular network. Principles of centrality measures were employed in the supply chain management to show that the weighted-ness of the edges/nodes together with the clique structure that emanates from it can be a pointer to centrality or otherwise of members of a group in the network of a distribution system. As expected, it was affirmed that the nodes belonging to the high clique members have a high percentage of being chosen/predicted as the most likely distribution centre. We examined the cliques of the weighted centrality matrix for the distributed system of a supply chain management network, and from the outcome we are able to predict a location of a new distribution centre in and around a particular area/region with an accuracy of more than 66%. In addition, the distinction between the notion of link-weightedness and node-weightedness were clarified.