作者: Paul Barford , Igor Canadi , Darja Krushevskaja , Qiang Ma , S. Muthukrishnan
关键词:
摘要: Over the past decade, advertising has emerged as primary source of revenue for many web sites and apps. In this paper we report a first-of-its-kind study that seeks to broadly understand features, mechanisms dynamics display on - i.e., Adscape. Our takes perspective users who are targets ads shown sites. We develop scalable crawling capability enables us gather details including creatives landing pages. strategy is focused maximizing number unique harvested. Of critical importance our recognition user's profile (i.e., browser cookies) can have significant impact which shown. deploy crawler over variety websites profiles yields 175K distinct ads. find while targeting widely used, there remain instances in delivered do not depend user profile; further, vary more than websites. also assess population advertisers seen identify 3.7K entities from business segments. Finally, when specific types generally correspond with profiles, users' patterns visit.