Maria Diapouli: Abstract

Targeted display advertising became one of the crucial topics in online advertising sector mainly due to the size of information that is collected regarding web users browsing activities. Billions of transactions every day involve millions of users, websites and advertisers. Advertising companies could improve their efficiency and competitive market position through data analysis and management.

In this presentation Crimtan an online advertising provider though a KTP collaboration with the University of Brighton is applying data mining and modelling techniques to deliver advertisements to web users who have no previously observed an advert but are good prospects, they are likely to take a desired action such as purchasing a product, known as conversion. Building user’s profiles requires the ability to estimate the probability that the user who sees the ad on his or her browser will take an action. However, conversion probability estimation is a challenging task since conversion event occurs rarely.

Decision tree algorithms are used to build users profiles utilizing past behavioural observations among user, with the aim to reveal the characteristics of a user who is likely to convert. Practical data mining considerations like data imbalance and testing various benchmark machine-learning algorithms were also presented along with the experiments results. The results are very promising and point out that there is a need for further studies in the area of using data mining in marketing.

Organised by BCS SGAI The Specialist Group on Artificial Intelligence http://www.bcs-sgai.org