A Hybrid Approach with Joint Use of Tag and Rating for Vehicle and Cargo Matching

This study develops and evaluates the joint use of improved methods to pair trucks with cargo on a vehicle-cargo matching (VCM) platform. The improved content-based VCM method recommends trucks to cargo owners based on their requirements. Meanwhile, a hybrid tag-based and collaborative-filtering (CF) approach accurately predicts shipper's ratings of drivers. Then the two methods are combined to obtain the final preference-based VCM (PVCM) model to make recommendations for shippers. The results of experiments indicate that proposed methods are competitive to typical cosine and CF methods in terms of Matching degree and Preference.