Customer-Perceived Value and Social Media Analytics: How Supplier Evaluation Can Benefit from Aspect-Based Sentiment Analysis and Fuzzy Inference

Abstract

Understanding customer-perceived value is essential for driving strategic supplier development and operational excellence in modern supply chains. As customer sentiments increasingly manifest in social media, natural language processing (NLP) techniques carry the potential of extracting actionable insights from unstructured data. This study proposes a novel decision-making model that combines aspect-based sentiment analysis (ABSA) with fuzzy inference systems (FIS) to support supplier evaluation based on customer value perception. Bridging symbolic and sub-symbolic AI, the model quantifies sentiment polarity, subjectivity, and aspect relevance from social media content, integrating this information into a multi-stage fuzzy logic framework for large-scale group decision-making (LSGDM). Unlike conventional supplier evaluation methods, this approach operationalizes concept-level affective information by fusing customer sentiment information with supply chain operational data. An illustrative application in the smartphone industry demonstrates the model’s ability to analyze social media data from the X platform and generate quantitative indicators reflecting customer value perception. The model effectively incorporates these insights into supplier evaluation, highlighting suppliers’ strengths and areas for improvement. The results show that sentiment-informed supplier assessment enables more responsive and customer-aligned development strategies. The proposed approach highlights the benefits of integrating sentic computing principles into supply chain analytics, showing the model’s capability to capture customer perceptions and make them a driver for continuous improvement initiatives in supplier development.