Content Filtering
When creating a Network Traffic Policy in the Traffic Policies screen, select the Content Filtering field and select any option from the dropdown menu according to the user’s need. The traffic will be restricted as per the chosen option.
What is Content Filtering?
Kognitive Networks utilizes WebTitan Cloud Services for Content Filtering Services [1]. The Current EdgeOS release only supports three Primary Content Filter Categories [2]:
Malicious
Subcategories: Criminal Skills/Hacking, Anonymizer, Compromised & Phishing/Fraud. Spam, Spyware & Malicious Sites.
Adult
Subcategories: Mature, Nudity and Pornography/Sex
Gambling
Online Gambling
The strength of the Kognitive Content Filtering solution is the Malicious protection, and it is recommended to always enable Malicious Content Filtering.
Malicious
Accuracy and reliability of the category updates & construction are noted here:
Coverage
WebTitan’s network of 500 million users ensures over 99.9% coverage of the active web traffic.
Accurate
A combination of machine learning and human quality assurance to maintain high accuracy, and it is updated in real-time.
Granular
53 different web categories in over 200 languages structed via WebTitan via the following framework.
Reporting is provided within the EdgeOS Usage Status, and Kognitive Cloud Service Portal Content Filter Reports. Malicious Category construction and capabilities are defined here:
Web Titan Filter provides leading-edge AI driven threat intelligence including:
Real-time URL threat protection powered by 650 million end users and growing.
Our real-time Database contains 3 million malicious URLs, phishing sites and IP addresses.
Every single day we identify 100,000 NEW malicious threats and URLs.
Real-Time Categorization and Malicious Detection for Domains, Full-Path URLs, and IPs
Crowd-sourced coverage and protection powered by over 650 million end users
Industry-leading 99.9% Coverage and over 99% Accuracy of ActiveWeb URLs & IPs
100% coverage of the Top 1 Million most visited websites (ActiveWeb)
Highly granular categories which include malicious, objectionable, and topic-based groups
10 Malicious categories including Botnet, Phishing/Fraud, Malware Distribution Point, Command and Control, Cryptocurrency Mining, Ad Fraud, and more.
Malicious URLs are revisited daily to determine if they are still infected, abandoned, or cleaned. This ensures that our malicious database stays as “fresh” and accurate as possible.
Support for over 200 languages
Real-time continuous updates provide up-to-the-minute protection against newly identified malicious sources
The malicious detection solution employs a crowd-sourced approach to obtaining a constant stream of URLs for analysis. The continuous stream of ActiveWeb URLs comes from our global network of over 650 million end users across a number of high traffic markets. We leverage an integrative, multi-vector approach to detect, monitor, and accurately categorize new threats as they occur. WebTitan uses AI-based services and human-supervised machine learning techniques and verification methods to achieve unmatched coverage (99.9%) and accuracy (99%+) for ActiveWeb domains, sub-domains, full-path URLs, and IP addresses. That includes 100% coverage of the Top 1 Million most visited sites and supports 200 languages, worldwide.
[1] https://zvelo.com/ is the classification technology utilized by WebTitan (and others such as Zscaler, Amazon)
[2] By Mid-2024 the EdgeOS 7 will support all 53 Content Categories.