Ubiquitous hotel recommendation is a highly popular type of location-aware service. However, existing recommendation systems have several problems. This paper proposes a fuzzy-weighted-average (FWA) and backpropagation-network (BPN) approach for overcoming the hindrances of ubiquitous hotel recommendation and improving its effectiveness, whereby FWA is applied to evaluate the overall performance of a hotel. A BPN was constructed to defuzzify the overall performance. In addition, the personally preferred index is proposed for addressing the traveler choices of a dominated hotel. The effectiveness of the proposed methodology was tested using a field study in a small region in Seatwen, Taichung City, Taiwan.