Abstract
Images are vitally important in interesting consumers and helping them to make decisions. Images of a hotel are particularly important and were used to sell hotels even before the Internet, when travel agencies would often have brochures about hotel properties that they used to entice travelers. On many online travel agency (OTA) websites, the hotel's image can take up 33% of the space on the hotel property page, but the importance of this image in the decision-making process has yet to be studied. For many OTAs, there are currently no quantitative analytic methods that help determine which image to display in this critical location. In this research, we use deep learning to extract information directly from hotel images and we apply image analytics to understand the importance of this information in the online hotel booking process. To provide managerial insights, we will combine a prediction model, with the t-distributed Stochastic Neighbor Embedding (t-SNE) to classify and understand the types of images hotels generally use as their thumbnail or "champion" image and what aspects of these images elicit consumers to consider and book a hotel.

This publication has 13 references indexed in Scilit: