Targeted advertisement through website user profiling

Abstract

An analysis of a travel portal, ChooseToTravel.com, showed that the travel portal general information section sees around 60% of the visits, the majority being referred directly, hence having most of the visitors visiting this section of the website not visiting products sections. These visits do not procure any income to the travel portal as visitors would be just browsing researching their destination and not viewing any products. In order to take advantage of these "lost" page visits, there need to be links in form of adverts or any other form that can attract the visitor's attention and target their interests so to redirect them from an information page to a product page. The proposed solution for this is the implementation of a profiling system that analyses the visitor's browsing pattern in real time and generates recommendations in the form of adverts or links to be displayed to the visitor accordingly. The system also intends improving the options available in the product pages so to recommend alternative products to the visitor in case the currently viewed one does not match their preferences. These recommendations aspire in turning "lookers" into "bookers". Through the creation of this system we intend to address how these information pages could also create a form of income. This shall be achieved through recommended navigation helping visitors find the information required more quickly, while also aiming at redirecting visitors to product pages in order to increase the chance of any particular visitor buying a product. These functionalities shall run without asking implicit information to the visitor, and gather all the required data explicitly without bothering the user with registrations, logins or filling up forms and risking getting them uninterested in using the system.