The process of selecting a supplier is a significant decision in supply chain management, as it can greatly impact the quality and cost of the procured products or services. This becomes even more important when shopping online, as there may be numerous options and thousands of reviews for a specific product type. In this research a novel hybrid methodology for supplier selection in e-commerce environment is introduced, which combines text mining and sentiment analysis of large customer review data and expert opinions of fuzzy multiple criteria decision-making (MCDM). Supplier selection requires expert perspective to determine the relevant criteria and assign them proper importance weights. Artificial intelligence is used to extract and interpret the emotional tone of customer reviews, adding valuable input to the determination of evaluation criteria and the rating of alternatives. The q-rung orthopair fuzzy set MCDM methodology, which is useful in situations with high levels of uncertainty or conflicting objectives and allows for the conversion of these qualitative expert opinions into a quantitative evaluation and determination of final criteria and their importance with the help of decision-makers' wisdom. By combining Artificial Intelligence techniques and MCDM approach, a more comprehensive and nuanced methodology to supplier selection is offered, taking into account both the qualitative and quantitative aspects of the decision. As two different real-life case studies, office chairs and robot vacuum cleaners from Amazon.com, both characterized by a substantial number of customer reviews and various features, were selected. Users' perspectives on multiple product features were identified, allowing for informed decisions and the provision of feedback on potential product improvements. Remarkably, the proposed methods aligned with the star ratings provided by 40,000 Amazon customers, underscoring the reliability and validity of the method. The proposed approach stands out in the supplier selection field with its innovative combination of sentiment analyses of customer review and perspectives of the decision experts, offering a cutting-edge tool for e-commerce managers to select or evaluate suppliers in e-commerce environment.

选择供应商的过程是供应链管理中的一个重要决策，因为它可以极大地影响采购产品或服务的质量和成本。在网上购物时，这一点变得更加重要，因为特定产品类型可能有多种选择和数千条评论。在这项研究中，介绍了一种新颖的电子商务环境中供应商选择的混合方法，该方法结合了文本挖掘和大型客户评论数据的情感分析以及模糊多标准决策（MCDM）的专家意见。供应商选择需要专家的视角来确定相关标准并为其分配适当的重要性权重。人工智能用于提取和解释客户评论的情绪基调，为评估标准的确定和替代方案的评级添加有价值的输入。q 梯级正对模糊集 MCDM 方法，在具有高度不确定性或目标冲突的情况下非常有用，并允许将这些定性专家意见转换为定量评估并在决策的帮助下确定最终标准及其重要性- 制造者的智慧。通过结合人工智能技术和 MCDM 方法，提供了一种更全面、更细致的供应商选择方法，同时考虑了决策的定性和定量方面。作为两个不同的现实案例研究，我们选择了亚马逊网站上的办公椅和机器人吸尘器，这两种产品都有大量的客户评论和各种功能。确定了用户对多个产品功能的看法，从而可以做出明智的决策并提供有关潜在产品改进的反馈。值得注意的是，所提出的方法与 40,000 名亚马逊客户提供的星级评级一致，强调了该方法的可靠性和有效性。所提出的方法以其对客户评论的情感分析和决策专家的观点的创新结合而在供应商选择领域脱颖而出，为电子商务管理者在电子商务环境中选择或评估供应商提供了尖端工具。