text classification, Word2Vec, deep learning, neural network, Web news, unsupervised learning.

The consequences of disruption to a supply chain can be damaging and costly. A firm's management needs to utilize its resources effectively to minimize the economic, and possibly environmental impact. This paper proposes the recovery model of a two stage serial supply chain subject to supply disruption. The system consists of a single manufacturer and a single retailer, subject to random supply disruption. The manufacturer keeps extra inventory as safety stock to be used at the time of disruption. In addition, the system may have stockouts in the form of partial backorders and lost sales. This study will incorporate the environmental effect consideration in the recovery model by including the carbon dioxide emission caused by the transportation activities during the recovery cycle. The objective of the optimization model is to determine (1) the new recovery schedule of the manufacturer and retailer, (2) the optimal quantity for safety stock, and (3) the carbon emission cost impact during recovery. The model developed is solved using LINGO, where numerical examples and sensitivity analysis are provided to test the feasibility of the model. The result of this study is an introduction of a recovery model that incorporates environmental consideration in the disruption recovery decision-making process. [ABSTRACT FROM AUTHOR]