An approach to situational market segmentation on on-line newspapers based on current tasks

We discuss how a task-based situational market segmentation may be applied to on-line newspapers, distinguishing between fact finding, information gathering and browsing. During a period of four weeks we had 41 users keep a diary and recorded their surfing behavior on different on-line newspapers. The results of a Naive Bayes classification with feature selection indicate that content-related attributes such as the number of news categories browsed are indispensable for task recognition.