Stacked Classification Framework
Spam Soap ensures industry-leading filtering accuracy using a multi-layered strategy that combines more than 20 forms of spam, virus, content, attachment, and email attack filtering technology.
The key to Spam Soap’s effectiveness is the Stacked Classification Framework® spam detection system, which is powered by patented technology and combines the industry’s most effective spam-fighting filters and techniques.
Through an aggregation and analysis of spam-likelihood scores, Spam Soap blocks over 99 percent of spam while maintaining industry-leading low “false-positive” rates (legitimate emails misidentified as spam). As new techniques and filters are developed, Spam Soap adds them to the Framework, further ensuring spam-filtering effectiveness.
Filtering specifically for spam, Spam Soap layers include:
- IP Reputation Connection Manager: This filter operates at the front of the Stacked Classification Framework and rates the reputation of every incoming message, based on IP reputation data that Spam Soap collects on an on-going basis. Connections are dropped for all messages that originate from IP addresses with a reputation for sending spam.
- Premium Anti-Spam Multi-Language: This filter provides Spam Soap with a global view of spam traffic, which enables the system to defend against real-time spam attacks and rapidly identify zero-hour spam, regardless of language. The filter also identifies image-based spam and phishing emails, and is continually updated based on real-time feedback from a global network of users.
- Statistical Filtering: Spam Soap’s probabilistic filtering uses a statistical Bayesian algorithm to determine the probability that an email message is spam, based on how often elements in that message have appeared in other spam emails.
- Sender Policy Framework (SPF)/Sender ID: For inbound messages, Spam Soap can check if the message has an associated SPF/Sender ID record. If there is a record, it can help determine if the email sender’s domain is included on a list of IP addresses authorized to send email from that domain.
- Industry Heuristics: Spam Soap incorporates thousands of successful industry-wide spam-fighting rules into its filtering layers.
- Proprietary Heuristics: Spam Soap experts write and update thousands of proprietary rules to block spam using real-time data from the Spam Soap Threat Center.
- Reputation Analysis: Reputation analysis votes on the probability that the message is spam, based on comprehensive information about the source of the message. The reputation of the sender is rated, based on the percentage of spam messages sent from that IP address in the past.
- URL filtering: URL filtering works by comparing embedded links found in email messages to URLs associated with identified spam.
- Reputation-based RBL filtering: Spam Soap assigns a level of trust to key real-time blackhole lists (RBL), which rates the reputation of the RBL based on its accuracy at blocking spam.
- Deep Content Analysis attachment filter: Unlike any other solution on the market, the Deep Content Analysis filter allows all email attachments to be treated holistically and therefore enables Spam Soap to analyze these messages in their entirety to determine if the email and/or the attachment contain spam or malware before they reach the customer’s network.
Spam Soap also integrates the following domain-level allow and deny lists and distributed deny lists into its comprehensive service to fight spam and other email threats:
- Domain-level allow and deny lists: Specifically designed to protect against spam, inappropriate content, and email attacks, domain-level black and white lists filter and block unsolicited messages.
- Distributed deny lists: Providing exceptional protection against spam, distributed black lists comprise a number of real-time subscription services including the Mail Abuse Prevention System (MAPS) and Spam Soap’s own global deny lists, which include multiple lists of known spammers and their IP addresses.
- Recipient deny lists (Address): This type of filtering is designed specifically to filter for content and relieve network servers from attempting repeatedly to deliver mail to invalid addresses.
- User-level black and white lists: Through regularly delivered Spam Soap Spam Quarantine Reports, end users have the flexibility to develop their own personal allow and deny lists.
