Incorporating elements from image recommender systems within a personalized virtual tour framework

A major vehicle that makes personalization possible is the recommender system that offers solution to overcome today’s world of information overflow. It provides specific suggestions based on user requirements, profiles or similar cases that have been previously handled by the system. This paper intends to illustrate the key elements summarized from previous image recommender systems, to be embedded within the suggested personalized virtual tour framework conducted in this research. This discussion will start by reviewing recommendation strategies that implement different ways of providing recommendations. As the focus of this paper is on image recommender systems, five key elements for better recommendations have been compiled based on a literature survey. Subsequently, these key elements were incorporated within a preliminary hybrid personalized virtual tour framework known as the ‘See What You Want, Feel What You See’ (SeeWYW, FeelWYS) model. Basically, the objective of this research on image recommendation is to implement a hybrid recommendation approach, which includes the socio-demographic and context-aware recommender engines. The socio-demographic recommender will deliver a suitable virtual route based on user demographic profiles. Following the suggested virtual route, a context-aware recommender will present a sequence of panorama that can be adapted based on user’s emotion as contextual information. It is hoped that these initial findings will provide insight on how to produce improved personalized and adaptive recommender system with good usability and good user feedback as well.