Spam & Fake Reviews
Spammers and fraudsters can easily circumvent rules engines or supervised machine learning solutions with sophisticated attack techniques. DataVisor’s Unsupervised Machine Learning Engine provides the next level of defense against these sophisticated techniques by analyzing the hidden connections between accounts.
How Spammers Evade Detection
Attackers simulate real user activity by uploading content scraped from other sites, fooling rules engines, supervised ML, and human reviewers.
Cyber-criminals employ armies of low-cost, on-demand human labor to bypass captchas and create realistic reviews and spam.
Fraudsters incubate fake accounts for months with simulated user activity to evade account-age and activity-based rules.
Fraudsters utilize proxies, VPNs or cloud-hosting services to hide their tracks from IP blacklists and rules-based systems.
Why UML Can Stop Sophisticated Social Media Abuse
DataVisor’s Unsupervised Machine Learning Engine is uniquely capable of stopping sophisticated, evolving social media abuse because it analyzes all accounts and events at once, uncovering the hidden connections between them. This approach allows DataVisor to detect spam or fake reviews without training data or labels, and even when individual accounts do not appear suspicious when viewed in isolation.
Detect Unknown Threats
Detect new and continually evolving attacks faster, without waiting for training data or labels.
Accuracy and Coverage
Detect malicious accounts at account registration before they can send spam or publish fake reviews.
Learn How DataVisor Stops Fake Reviews & Spam
Srigurunath Chakravarthi, Head of Trust at Pinterest discusses how DataVisor has helped detect more spammers faster with Unsupervised Machine Learning.
Watch this recording of a webinar hosted by Julian Wong, Technical Architect at DataVisor, and his special guest, Jim Blomo, Director of Engineering at Yelp, to learn how Yelp stops fake users while connecting people with local businesses.
The DataVisor Platform
Unsupervised Machine Learning Engine
Supervised Machine Learning Engine
Automated Rules Engine
Global Intelligence Network
What’s Happening with Spam & Fake Reviews
Fraudsters are constantly coming up with new and innovative ways to commit fraud. Today we are taking a look at product listing fraud, a relatively new type of fraud that is a rapidly growing problem for online marketplaces.
A successful spam campaign is one that obtains maximum return-on-investment (ROI) to the spammer. This means that a spam campaign must reach as many end users as possible, must be robust in the face of blacklisting efforts, and must be scalable. This blog post describes some of the recent techniques employed by spammers to distribute malicious URLs on social media platforms as observed by DataVisor.
The DataVisor Online Fraud Report took a look at our base of more than one billion users across 172+ countries in the world. Using this massive amount of data, we were able to identify some of the favorite tools and attack techniques that online criminals from around the globe favor when doing their dirty work.