This paper proposes a highly intelligent digital signage system based on the IoT (Internet of Things) edge equipped with a learned advertisement recommendation model. The proposed system consists of a server and an edge. The server manages the data, learns the advertisement recommendation model, and the edge determines the advertisement to be promoted in real time using the learned advertisement recommendation model. The ad recommendation model consists of a selection of products and a prediction of their purchasing probabilities. In the screening phase, the user-based information and product metadata vectored into DNN are entered to derive the product that is worth purchasing. We use a soft max function to predict the purchase probability of selected goods. Finally, the most suitable advertisement is selected by using the predicted purchase probability of the community. The proposed system does not communicate with the server. Therefore, decides the advertisement with the learned model on the edge. This also applies to digital signage that requires immediate response to many users.
CITATION STYLE
Lee, K., & Moon, N. (2021). Intelligent Digital Signage Using Deep Learning Based Recommendation System in Edge Environment. In Lecture Notes in Electrical Engineering (Vol. 715, pp. 37–43). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-15-9343-7_6
Mendeley helps you to discover research relevant for your work.