Unscramble social media power for waste management: A multilayer deep learning approach

Abstract

In an industry 4.0 era, manufacturers must make the right decisions concerning reutilizing and recycling returned goods to lower waste by connecting and integrating societal behaviour with industrial practices. Since customers would not buy unsatisfactory goods, managers have to decide whether to reuse or recycle them. Proposing a framework in which consumers involve in supply chain managerial decisions leads to customer-centric reverse logistics by introducing a Decentralized consensus decision-making concept. Moreover, an industry benchmark would aid managers and policymakers. Based on Logistics 4.0, this study developed a framework and used social media data to improve reverse logistics decision-making using artificial intelligence - a deep learning hybrid method that combines convolutional neural networks and long short-term memory networks. Furthermore, this study shows how to use sentiment analysis algorithms to analyze positive and negative feedback from customers and develop high-efficiency decentralized disposition decision-making techniques for reverse logistic managers. Without the need for human intervention, corporations may use the framework to smart scrutinize their customers' feedback and sentiments to make strategic reverse logistics decisions for reducing returned goods, waste, inventory levels, and costs while increasing productivity, benefits, supply chain sustainability, and customer loyalty, resulting in a more effective movement towards green logistics, a competitive advantage, and increased profits for the corporation. While the framework is flexible enough to be utilized in various fields, including electronics and automobiles, it significantly reduces the risk of biased feedback resulting from not considering a particular language or region.