Modeling and Analyzing User Contexts for Mobile Advertising

Abstract

Context-aware advertising is one of the most critical components in the Internet ecosystem today because most WWW publisher revenue highly depends on the relevance of the displayed advertisement to the context of the user interaction. Existing research work focuses on analyzing either the content of the web page or the keywords of the user search. However, there are limitations of these works when being extended into mobile computing domain, where mobile devices can provide versatile contexts, such as locations, weather, device capability, and user activities. These contexts should be well categorized and utilized for online advertising to gain better user experience and reaction. This paper examines the aforementioned limitations of the existing works in context-aware advertising when being applied for mobile platforms. A mobile advertising system is proposed, using location tracking and context awareness to provide targeted and meaningful advertisement to the customers on mobile devices. The three main modules of this comprehensive mobile advertising system are discussed, including advisement selection, advertisement presentation, and user context databases. A software prototype that is developed to conduct the case studies and validate this approach is presented.