VISIO: a visual approach for singularity detection in recommendation systems

Reviews are a powerful decision-making tool for potential new customers, since they can significantly influence consumer purchase decisions, hence resulting in financial gains or losses for businesses. In striving for trustworthy review systems, validating reviews that could negatively or positively bias new customers is of utmost importance. To this goal, we propose VISIO: a visualization based representation of reviews that enables quick analysis and elicitation of interesting patterns and singularities. In fact, VISIO is meant to amplify cognition, supporting the process of singling out those reviews that require further analysis. VISIO is based on a theoretically sound approach, while its effectiveness and viability is demonstrated applying it to real data extracted from Tripadvisor and Booking.com.