BOX Solutions Corp.

Most current Anti-Spam technologies can be categorized as “Content Filtering” technologies. These outdated technologies receive and scan every message in its entirety, as in the MUA layer, to be cross-matched against keyword databases in order to make judgments according to various analysis algorithms, such as statistical weighting or pattern comparison. Thus, from the point of view of the core technologies, these various message content analysis methodologies are not fundamentally different from one another - such methodologies are often known as “Content Filtering “, “Database”, “Blacklist Collections and Analysis”, “Recipient Control”, and various statistical strategies. However, the common feature of these mail content analysis technologies is that it is notoriously easy for such engines to misjudge whether a message is Spam or non-Spam. That’s why none of the above technologies are decisive-judgement technologies. Furthermore, an even more fundamental flaw in such technologies is that they are deeply dependent on language semantics or database mechanisms and must be frequently updated and refreshed in order to maintain consistent judgement quality.

SpamTrap judges messages according to communication behavior only; thus it is specific and deterministic.

Instead of scanning mail content, SpamTrap analyzes communication behavior (Transmission Data) to make judgements. SpamTrap is the security guard standing in front of the enterprise gateway: it collects and traces transmission data only (beginning from transmission handshaking to “mail subject”) then it can judge if the message follows Spam transmission patterns. SpamTrap follows IRTF/IETF/RFC standards; that is, SpamTrap follows behaviour policy instead of content policy.

SpamTrap spots spam behavior as if it were a human security guard spotting an intruder using a fake ID, breaking a window, sneaking in, wearing a fake/stolen company uniform, refusing to sign in, ... etc.

SpamTrap’s Key features:

SpamTrap specifically analyzes truth/false values of Message Transmission Data in MTA layer.
SpamTrap scans no mail content, scan-base is much smaller and makes it highly efficient.
SpamTrap is language-independent; Spam Filtering is effective for all countries and languages.
SpamTrap default Communication Spam Behaviour Patterns make easy management and high performance possible. No patterns or databases need frequent updates.
With a token-based algorithm, SpamTrap executes filtering & judgement in the mail transmission and delivery process.
All policies of SpamTrap are executed as an efficient, single operation process.
Analyzing Mail Transmission Data (in size of 3~10K only), with dynamically generated code and no reliance on database, SpamTrap has extremely high system efficiency and stability.