NETCOM Research Group

Researchers from IMDEA Networks, UC3M, NEC Labs Europe and Politecnico of Torino have reached striking results in their recently published paper Understanding the detection of fake view fraud in Video Content Portals. For the first time, it has been demonstrated that video portals are very vulnerable to fraud by fake views from robots (bots).

IMDEA Networks’ researchers have recently presented their research results on one of the hottest topics in today’s Internet: the business behind advertising. This business is fueling today’s Internet, as it is the main economical driver behind the network.

In the recently published paper Understanding the detection of fake view fraud in Video Content Portals, the research team formed by IMDEA Networks, University Carlos III of Madrid, NEC Labs and Politecnico of Torino have shown that online video portals are charging marketers for advertisements placed on YouTube even when the actual viewer of the ad is in reality a robot rather than a human being.

To perform the experiment of testing the performance of fake view detection systems, the researchers applied measurement tools developed specifically for this specific purpose to the five major online video portals. The research results revealed that YouTube's detection system significantly outperforms all the others. Despite this, the systematic evaluation that was carried out has shown that it may still be susceptible to simple attacks. Furthermore, they found that YouTube penalizes its users' public view more aggressively than the monetized views "(the ones for which advertisers are charged). In practice, this means that views identified as fake and discounted from the public view-counter are still monetized.

These results have led to speculation on whether enough is being done by online media magnates such as YouTube and Google to avoid advertisers wasting their investments on fake views of their online commercials. Despite their policy to compensate users financially when attacks are discovered, this practice still places the burden of the risk on the advertisers, who pay to get their ads displayed.

Arturo Azcorra, Director of IMDEA Networks and Albert Banchs, Deputy Director have participated in this research effort, which has been led by Rubén Cuevas, from University Carlos III of Madrid. The research results prove that video portals are substantially vulnerable to fraud by fake views from robots. YouTube which is the industry leader and the best online video portals site included in the experiment at detecting fake views, is still substantially vulnerable. YouTube, which is the industry leader and the online video portal with the best results in the experiment at detecting fake views, is still substantially vulnerable.

The results of this research effort represent a breakthrough in the understanding of the vulnerability of the main source of income on the Internet, and have had a very profound impact on the main players behind the Internet business. This had led to a significant number of well-established international media such as the Financial Times, BBC, The Guardian, The Times (UK), the Brasilian O Globo and the Italian Corrier della Sera publishing the story. Also national big players such as El País and EuropaPass followed suit and published the research results, emphasizing their importance for society as a whole and the advertising industry in particular. As regards the Financial Times – the first newspaper to publish the story – the article was the second most viewed on the day of publication.

Following the first wave of the significant media coverage, major industrial players have reacted promptly and expressed their concern after the revelation of the research findings. The Financial Times (among others) followed up with a second article in which they request Google to address the issue of fraud, and thus, make more of an effort to protect marketers from deceptive fake views of their advertising.