A Novel Feature Engineering Framework in Digital Advertising Platform

摘要

Digital advertising is growing massively all over the world, and, nowadays, is the best way to reachpotential customers, where they spend the vast majority of their time on the Internet. While anadvertisement is an announcement online about something such as a product or service, predicting theprobability that a user do any action on the ads, is critical to many web applications. Due to over billionsdaily active users, and millions daily active advertisers, a typical model should provide predictions onbillions events per day. So, the main challenge lies in the large design space to address issues of scale,where we need to rely on a subset of well-designed features. In this paper, we propose a novel featureengineering framework, specialized in feature selection using the efficient statistical approaches, whichsignificantly outperform the state-of-the-art ones. To justify our claim, a large dataset of a runningmarketing campaign is used to evaluate the efficiency of the proposed approaches, where the resultsillustrate their benefits.