Promoting reverse logistics decisions using a new hybrid model based on deep learning and failure mode and effects analysis approaches

The problem of reusing and recycling the returned products plays a crucial role in mitigating waste. Therefore, authorities must make the best decision in such situations. However, this problem is a paradoxical decision because different components often conflict with each other, which can impact the decision-making process. The proposed framework uses sentiment analysis algorithms to help decision-makers adopt the best reverse logistics decision strategy based on customer feedback. The framework provides a procedure for extracting, categorizing, and analyzing customer opinions. It strategically decides in reverse logistics to increase profit, efficiency, and customer satisfaction while reducing the returned products, costs, and waste. The framework has a high potential for utilization in a wide range of industries, so the probability of a biased opinion resulting from the limitation of taking into account a specific location or time is significantly diminished. This paper employs a big data mining approach to optimize the decision procedure in reverse logistics by using social media data based on customer satisfaction. To demonstrate the capability and effectiveness of the proposed framework, a real case study based on the Apple Notebook, a branch of the electronics industry, is illustrated. Consequently, a separate sentiment analysis based on a recurrent neural network (RNN), a deep learning approach, is fulfilled for notebook features and models. The framework can determine the most appropriate disposition decision in reverse logistics. Furthermore, a failure mode and effects analysis (FMEA) procedure was employed to make some suggestions about Apple.