A Framework to Improve the Confidence of Restaurants User Reviews Based on Facial Analysis

The quality reviews from users impact significantly restaurants' reputation and their economic success. Their customers are influenced by other reviews of their experience with restaurant service and therefore, they tend to prefer restaurants already classified as excellent. Unfortunately, a significant number of users' reviews are biased and non-trusted. These reviews misguide other users and lead to unexpected, sub-par experiences, leaving them frustrated and disappointed. To overcome this scenario, this work proposes a framework to improve the confidence of restaurant users' reviews at scale. For that purpose, user ratings are evaluated if they follow their facial emotions at the moment they are submitting their reviews. The experimental results denote the feasibility of the proposed framework to improve the confidence of restaurant users' reviews. Decomposing their components into loosely coupled services helps to manage the complexity and improve scalability, performance, and agility. The framework performance and scalability allow a large adoption and integration with third-party users' review solutions.