An Empirical Framework for Recommendation-based Location Services Using Deep Learning

Abstract

The large amount of possible online services throws a significant load on the users' service selection decision-making procedure. Α number of intelligent suggestion systems have been created in order to lower the excessive decision-making expense. Taking this into consideration, aν RLSD (Recommendation-based Location Services using Deep Learning) model is proposed in this paper. Alongside robustness, this research considers the geographic interface between the client and the service. The suggested model blends a Multi-Layer-Perceptron (MLP) with a similarity Adaptive Corrector (AC), which is meant to detect high-dimensional and non-linear connections, as well as the location correlations amongst client and services. This not only improves recommendation results but also considerably reduces difficulties due to data sparseness. As a result, the proposed RLSD has strong flexibility and is extensible when it comes to leveraging context data like location.