Grammar Structure Delimiters
Key word searches typically use simple Boolean statements to determine whether an email contains sets of concepts, wordsm or phrase that often appear together in content or spam, such as "toll" and "free." However, many emails may contain suspect sets of words without being a content or compliance violation, or spam, such as:
I traveled across the toll bridge today on my way to work. The change machine was broken so I received a free pass on the bridge.
Location and appearance also play an important role in determining accuracy. For instance;
the word "confidential", or a number of its equivalences like "classified" and "secret"
if appearing multiple times scattered throughout an email or attachment does not necessarily mean that the email is confidential, however if these concepts were located in the header, footer, title page, or summary information of a document, then they would definitely imply that the message and its attachments were of a confidential nature.
SecurExchange's Intelligent Content Analysis (ICA) reduces the number of false positives and increases the accuracy by searching for concept patterns by grammar constructs;
- sentence
- paragraph
- line
- bullet point
- centered text
and attachment layout;
- header
- footer
- title page
- table of contents
- index
- summary information
|
