Previous scientific studies as well as consulting firms have developed numerous Online Reputation Indices (ORIs), i.e. custom-tailored metrics intended to measure the emotions that people express towards a brand, product, or service in social media. These ORIs can provide useful information to assess the impact of marketing campaigns, social approval, and viral behavior of news and memes, among others. However, traditional ORIs are isolated metrics; thus, they are not suitable for determining the relative preference between two alternatives; e.g.: twice the number of “likes” does not imply that a given product is preferred two times as much as its competitor. This is an important constraint for ORIs, because stakeholders are driven to make relative and qualitative comparisons to weigh the alternatives’ value. Furthermore, relative comparisons are crucial for the systematic evaluation of alternatives in social decision making processes. The aim of this paper is to present a novel qualitative online reputation metric, a Comparison Matrix Geometric Index (CMGI), that considers both the accumulated emotions and the direct comparisons expressed in social media communications to infer the relative preferences among a set of alternatives.

先前的科学研究和咨询公司已经开发了许多在线声誉指数（ORIs），即用于量身定制的度量标准，旨在衡量人们对社交媒体中的品牌，产品或服务表达的情感。这些ORI可以提供有用的信息，以评估营销活动，社会认可以及新闻和模因的病毒行为等的影响。但是，传统的ORI是孤立的指标。因此，它们不适合确定两个替代方案之间的相对偏好；例如：“赞”次数的两倍并不意味着特定产品是首选是其竞争对手的两倍。这是对ORIs的重要限制，因为利益相关者被驱使进行相对和定性的比较，以权衡替代方案的价值。此外，相对比较对于社会决策过程中替代方案的系统评估至关重要。本文的目的是提出一种新颖的定性在线声誉指标，即比较矩阵几何指数（CMGI），该指标同时考虑了累积的情感和社交媒体通信中表达的直接比较，以推断出一组备选方案之间的相对偏好。