Abstract: Mobile devices are widely used todays with a huge number of applications usage that support users agreements. Rating matrices represent the users’ behaviors through the case study timeline and have some limitations like high dimensionality. Here, Singular Values Decomposition (SVD) is used to reduce feature space dimensionalities but with developed techniques. The paper proposes 3D-SVDs algorithm which splits the tracks and although the inducted rating matrices into multi-level data segments each one represents one period of time slices within system data. The extracted latent features from each level of 3D-SVDs are used to checking user similarities to his neighbors and then the system picks the group of nearest users to recommend their similar preferences to the current user. The system uses most frequent item recommendation technique to recommend best positions to the user from the latent grouped nearest user’s preferences in addition to using current user place from GPS as a new combination function to enhance recommendation list. The system recommends the daily preference places for users. Finally the results are shown on map supported application. The proposed system is built by using C#.NET and ArcMap GIS oriented programming for desktop version and Android Java for mobile devices version of the same system.