A Recommendation Framework For Real Time Personalization Route Recommendation Services

Recommendation systems have recently undergone rapid development in the field of information systems (IS). However, issues in existing recommendation systems may prevent their further development and application. These issues include privacy and rating, as well as the non-real time and inflexible nature of current recommendation systems.Therefore, this dissertation presents a recommendation framework to address the aforementioned issues and to extend the current recommendation systems to new fields and applications in IS. In addition, a prototype is established to illustrate the performance of the proposed recommendation framework, which can provide real-time, personalized route recommendation services to driving tourists based on vehicle-to-vehicle communication systems (V2VCS).The first section presents the recommendation framework, which is designed to address some issues of current recommendation systems as well as to extend the applications of conventional recommendation systems. The proposed recommendation framework consists of three main modules. The first module, the data process module, collects real-time item and personalization information for item recommendation. The second module, the item score module, ranks and rates the candidate items based on item attributes and user personalization preferences. The third module, the item generation module, explores anappropriate item in terms of user personalization requirements and real-time contextual information.The second section deals with the personalized route recommendation system in terms of the proposed recommendation framework, which offers a real-time, personalized route recommendation service to driving tourists based on V2VCS. The proposed recommendation service includes three main steps:routing information, employing of the fuzzy logic and multiple-criteria decision-making (MCDM) methods, and exploiting a generic algorithm method. In the route information step, personal requirement information and real-time traffic information are collected by V2VCS. In the second step, the fuzzy logic and MCDM methods are employed to rank and rate candidate routes in terms of route attributes and personalization preferences. The third step, which employs a genetic algorithm (GA) method, generates an optimal route for recommendation based on real-time contextual information and personalization requirements of driving tourists.The major contribution of this dissertation is that the proposed recommendation framework can solve some issues or problems associated with existing recommendation systems. In addition, this work can extend the applications of current recommendation systems to the provision of real-time personalized recommendation services.In summary, this dissertation covers recommendation systems, personalized systems, route recommendation systems, and V2VCS in the field of IS. This work sheds light on the provision of personalized recommendation systems in terms of user personalization, item information, and real-time contextual information.