If you’ve glanced at a headline in the last two years, then you probably know that artificial intelligence (AI) is transforming every facet of our personal and professional lives–machines that understand text, speech, and sense the environment around us, virtual agents and robots that change the way we communicate and work, autonomous vehicles that replace human-driven ones–and digital advertising is no exception. But behind the trade publication headlines about “the rise of the machines” and the ledes making the same tired Terminator references (guilty), the industry is still struggling to understand the implications of this new wave of disruptive technologies.
This complexity, and surely some of the industry’s confusion, is driven by the simple fact that we’ve been using terms like AI and machine learning interchangeably when they have vastly different meanings. So before we dig any further into what this technology can do to help fight back against fraud, we should try to get our definitions straight.
Artificial intelligence has existed in one form or another for decades, but recent algorithmic advances, the exponential increase of processing power and storage, and the explosion in data generated by consumers’ online activity has brought this technology to the mainstream. But AI itself is a blanket term. It refers to a number of big, potentially world-changing technologies that you may have heard of, like natural language processing, virtual agents, robotics, autonomous vehicles, computer vision, and machine learning.
By contrast, machine learning refers to analytic techniques that “learn” patterns in datasets. Machine learning allows computers to crunch data and discover these patterns in huge datasets that would be imperceptible to human analysts. The ad tech space is already reaping the benefits of machine learning in the fight against fraud. Machine learning techniques are extremely effective in combing through the millions of impressions generated by digital advertising campaigns and detecting patterns in these streaming data sets that are indicative of fraudulent activity, such as non-human traffic or other bot driven automated activity.
In the case of IAS, machine learning powers two critical pieces of our fraud detection technology.
Behavioral and network analysis – We use big data to distinguish real user behavior from bot behavior by looking at anomalies within site visitation patterns. Cohorts of bots tend to visit the same cluster of domains over and over because their behavior is automated. Detecting these patterns can allow us to surface bots based on their behavior. After all, most humans don’t visit the same sites, in the exact same order, multiple times per day. If these cohorts have only visited specific domains that can indicate a pocket of bot activity, we track these patterns and mark this traffic as fraudulent. Of course, machine learning techniques can also identify patterns in traffic that aren’t immediately obvious to human analysts.
IAS observes up to 10 billion impressions per day, the definition of big data. That scale allows us to build machine-learning models that can predict fraud which our peers with more limited scale and singular methodologies cannot. These models allow us to react quickly to new fraud innovations and be more resilient to bots that are trying to thwart our technology. Major consumer brands, like Uber and Amazon, are leveraging big data and machine learning to power major technological innovations, from driverless cars to drone delivery services, and we believe that marketers deserve access to the same predictive technologies to protect their digital investment.
Browser and device analysis – Machine learning allows us to identify invalid traffic sources by matching browser features to the user agent. While this type of determination is often mistakenly labeled deterministic, it would be impossible without employing machine learning methodologies to detect patterns within large data sets. Applied correctly, and powered by sufficient data, this method of detection can help to weed out entire bot networks.
Machine learning is at the heart of our current and future solutions for fighting fraud as well as providing brand safety protection. As a verification provider, IAS will continue to expand our efforts in two core areas.
AI-driven automation – Using AI to automate routine processes currently helps us to optimize high volume, rules-based work. We will continue to expand the use of AI automation in order to enhance scale for our solutions. This technology also helps to integrate cognitive capabilities into business flows in order to better analyze a given situation, learn from it, and reason through it without intervention from a human analyst.
Next-generation machine learning – We are continuing to expand our machine learning capabilities, especially through the adoption of deep learning and reinforcement-learning techniques based on neural networks. These two technologies depend less on existing domain knowledge and ensure optimized results through self-correction of the fraud prevention and detection models.
Ultimately, artificial intelligence will continue to drive change throughout industry and society. For digital advertising, much of that transformation will be powered by AI-driven automation and machine learning that can help us to grapple with the massive data sets produced by large, fully scaled, digital campaigns. IAS is committed to remaining on the forefront of this new technological revolution and to making sure that our clients and the industry remain educated about the potential of these technologies to fight fraud and help savvy marketers reclaim lost ad spend.
To learn more about our fraud solutions click here