What Are Driving Users to Click Ads ? User Habit , Attitude , and Commercial Intention

With the rapid growth of World Wide Web (WWW), online advertising is attracting more and more attention owing to its huge market potentials. In fact, in general sponsored search, the problem that what drive users to click ads is still underexplored in Web data mining, which essentially prevents us from further exploiting this market. In this paper, we propose to learn the factors that drive users to click ads by a large scale data driven study, on the basis of the insights researched by psychological studies which have been taken to understand user intentions. Three psychological factors are reformulated and verified in this work. They are user habit, attitude and commercial intention. We first introduce how to extract features according the approximate effect of these three psychological factors from the click-through log data of a commercial search engine. Then, we analyze the correlation between users’ ad click behaviors and those factors. After that, as an application, we build a prediction model integrating these factors to predict the probability that whether the user will click ads. The result of experiments on real user ad click log of a commercial search engine shows the prediction model can help classical machine learning algorithms perform better.