Deep Analysis of Visual Product Reviews

With the recent shift toward the digital world, online customers are increasing rapidly. To fulfill the market demand, competitive market strategies are introduced in the e-commerce industry. Analyzing customer feedback is becoming one of the integral parts of this market strategy, and indispensable to a service provider. In recent days, it can be noticed that customers upload purchased product images with their review scores. In this article, we undertake the task of analyzing such visual reviews, which is very new of its kind. In the past, the researchers worked on analyzing language feedback. However, here, we do not take any assistance from linguistic reviews that may be absent, since a recent trend can be observed where customers prefer to quickly upload visual feedback through a smartphone instead of typing language feedback. We propose a hierarchical architecture, where the higher-level model engages in product categorization, and the lower-level model pays attention to predicting the review score from a customer-provided product image. We generated a database by procuring real visual product reviews, which was quite challenging. Our architecture obtained some promising results by performing extensive experiments on the employed database. The proposed hierarchical architecture attained a 26.26% more accuracy than the single-level best comparable architecture.