Real-time prediction of information search channel using data mining techniques

One of the biggest challenges associated with using the Internet as a real-time marketing vehicle concerns digital media fragmentation. The vast amount of potential media sources and platforms that are available to marketers entails that it can be very difficult to formulate a succinct strategy through which they can interact with the existing and potential customers. While a large number of businesses typically use consumer's previous actions as a means of understanding their information search behavior, very little is understood about how demographics influence information search and the use of various digital platforms. Previous research has focused on identifying these demographic factors but, as yet, no one has developed a real-time model that is capable of predicting consumer's information search preferences. This research considers four information search channels: personal, marketer-dominated, neutral and experiential channels, and assesses the extent to which existing classification techniques, such as classification and regression tree, neural networks and support vector machines, can be effectively employed to forecast individual's search preferences according to their demographic context. It is envisaged that the development of a method that can accurately forecast search behavior will help organizations to ensure that they allocate marketing resources in an efficient and effective manner.