Story image

Appier fights ad fraud with the help of artificial intelligence

12 Dec 2017

Appier has managed to demonstrate AI’s ability to fight ad fraud by using a model that is able to pick out patterns that traditional models find it hard to detect.

"Ad fraud has become a major threat to the online advertising industry and is projected to cost advertisers billions of dollars," comments Appier chief technology officer Joe Su.

The company based its AI model on four billion data points that included ad clicks and app installs over a four-month period.

Its report found that the AI-based fraud detection was able to identity twice as many fraudulent transactions as a traditional model that uses rules.

One example is ‘the chameleon’, an instance where dishonest publishers disguise themselves as legitimate ones but then distribute fraudulent installs.

Another is what Appier’s AI terms ‘inventory burst’, in which a fraudulent publisher will generate abnormally high inventory count in the absence of an appropriate level of in-app registration activity.

"Traditional rule-based methods of detecting and mitigating ad fraud has its limitations. Appier believes an AI-based model is far more effective and we are seeing the benefits of that approach in just four months of analysis from our network,” Su continues.

Traditional rule-based models typically only look at one to three dimensions and generally work around rules developed by humans – which can work for simple and known ad fraud patterns, Appier says.

However an AI-based model is able to examine data across 80 dimensions, self-learn and is able to detect suspicious patterns that were previously invisible.

"Just as with cyber fraud or financial fraud, ad fraud is becoming more sophisticated and is constantly evolving, so it's important to be able to quickly identify and mitigate new threats. A traditional rule-based approach just cannot keep up with the fraudsters, an AI-based model is required to stay on top of evolving ad fraud patterns," Su concludes.

Appier chief data scientist Dr Hsuan-Tien Lin believes that AI will ‘delight and amaze’, according to a blog.

“We will change the way we live and work due to AI, and for the better. While AI systems can do many tasks better than humans and will take over repetitive, time-consuming or physically-dangerous tasks from humans, they are unlikely to replace humans altogether. The new AI-based applications will raise our quality of life and allow us to have more time to do what we want,” he says.

“In the Asia-Pacific region, businesses need to start thinking seriously about AI. Whether it is training an adaptable workforce, asking if the technology they want to purchase has AI components, and looking at building their own AI capabilities in-house, getting ready for an AI-capable world will stand them in good stead."

Salesforce continues to stumble after critical outage
“To all of our Salesforce customers, please be aware that we are experiencing a major issue with our service and apologise for the impact it is having on you."
D-Link hooks up with Alexa and Assistant with new smart camera
The new camera is designed for outdoor use within a wireless smart home network.
Slack users urged to update to prevent security vulnerability
Businesses that use popular messaging platform Slack are being urged to update their Slack for Windows to version 3.4.0 immediately.
Secureworks Magic Quadrant Leader for Security Services
This is the 11th time Secureworks has been positioned as a Leader in the Gartner Magic Quadrant for Managed Security Services, Worldwide.
Google puts Huawei on the Android naughty list
Google has apparently suspended Huawei’s licence to use the full Android platform, according to media reports.
Using data science to improve threat prevention
With a large amount of good quality data and strong algorithms, companies can develop highly effective protective measures.
General staff don’t get tech jargon - expert says time to ditch it
There's a serious gap between IT pros and general staff, and this expert says it's on the people in IT to bridge it.
ZombieLoad: Another batch of flaws affect Intel chips
“This flaw can be weaponised in highly targeted attacks that would normally require system-wide privileges or a complete subversion of the operating system."