Associated Services

ESP Data Loss Prevention and Flow Control

ESP™ (Enterprise Specific Protection) DLP is an advanced Email Data security and Control system that operates transparently on all inbound and outbound Enterprise Email.

It uses the power of classification to provide “always-on” deep content inspection and active, adaptive protection.

ESP™ distinguishes Firstwave Email security from every other competitor, providing protection of information and active risk mitigation. It is based on world class Australian patented technology and provides incomparable control for email and other messaging solutions.

Auto Classification

Often where messages are being submitted to EDRM’s or other control systems it relies on manual classification by all message recipients. This means that data classification is typically inconsistent or worse not classified at all either because it is too hard or too time consuming.

In most cases the underlying classification ontology is a corporate specific functional hierarchy (as opposed to a content based hierarchy. It may require the end user to navigate and choose from as many as 80,000 categories in order to find the appropriate classification for the document.

Inaccurate or missing classifications have the effect of lowering the overall quality of the EDRMS data repository and lowering its usefulness for topical document discovery and control.

With the addition of automatic classification technology, as each document is added to the repository an opportunity exists for an automatically generated shortlist of category recommendations, with associated confidence levels, to be presented to the user as part of the submission (setup?) screen. In this way, the end user still makes the final category choice, but is presented with a machine generated shortlist of choices based on an algorithmic analysis of the content of the new document being submitted to the system.

Automatic classification works by learning from past examples of how similar documents were manually classified into the repository. With each recommendation and subsequent manual ‘accept’ or ‘correction’ by the human end user, the machine learning algorithm is able to continue to learn and improve the accuracy of subsequent recommendations.

 

Back to Message Classification

For more information contact us......