ISheriff - Enterprise Content Security Industry Leader for Unified Content Appliances and Threat management software solutions
iSheriffTM SCM
iSheriffTM ITM
iSheriffTM ETM
iSheriffTM UCF
iSheriffTM Reporter
Key Features / Benefits
Deployment
Technical Specifications
Technology Overview
Real Time Behavioural Analysis
spacer
Download a free 30 day trial.

Contact us for more info.
see more >>
Products /
Real Time Behavioural Analysis
iSheriffTM Real Time Behavioural Analysis - Innovative Approach to Filtering

Our threat management solutions are distinguished from many available filtering products in that, when a URL is encountered which has not previously been definitively categorised, real time analysis of the indicative attributes of the actual data content is invoked as it passes through the filtering engine.

Before attempting to model a site, it must first have statistics gathered for it. This statistics gathering process is conducted in isolated threads ensuring no performance degradation occurs. As objects for a particular site are requested, a tree of the directory structure for that site is built. Statistical information is placed at each tree node for future analysis. A similar tree is also built that traces the objects that a client has retrieved. These two trees allow either the site content or client behaviour to be modelled.

The statistics collection thread monitors its resource consumption to prevent any adverse impact on performance. The senescence of statistical information is checked against configurable time-to-live parameters enabling removal of expired statistical data. Other configurable parameters include the minimum and maximum number of statistical threads to allow running, a threshold number of objects to be processed before additional threads are started and a number of objects at which the statistics gathering process is suspended.

Indicative attributes examined include relevant ratings of PICS self-rating systems, combinations of indicative words and phrases in the URL domain name and search path, combinations of indicative words and phrases in the header tags of the HTML source document and links out of the website to sites of known categorisation. Use of multiple real time modelling settings is fully configurable by the administrator for each group of users. The iSheriffTM models are highly effective in deriving content categorisations for websites of unknown categories.

The business benefit of this technology is very real. Imagine the scenario of an employee working to a tight deadline who needs to access data over the web as a matter of urgency to complete the required work, in today's world of ever changing content and new websites. If not already definitively categorised, the website is examined, a process taking only milliseconds and the site is then categorised, subject to human review. In the event that the audit result satisfies the parameters set by the system administrator, the staff member will have access to the critical information without any interruption to their work. Should the site not satisfy the predetermined parameters then there is very little doubt that the site is questionable and would be putting the organisation at risk should access be granted?

The real issue here is what would happen in this scenario if the iSheriffTM Media Control Engine were not available. In the event that the organisation has made the responsible decision to implement an internet management solution, they will be faced with a major dilemma - do they give staff unrestricted access to all sites not recognised by the latest "black list" or do they simply deny access completely. Both options have significant downsides. Do they risk severely hampering employee productivity or do they risk exposing themselves to legal penalties?

The real benefits of the iSheriffTM Media Control Engine are the fact that it allows an organisation to allow itıs staff access to work related sites while blocking all sites that may be undesirable, illegal or offensive and not just those recognised by the latest 'black list'.

The key differentiator lies in the fact that any solution based on lists that are downloaded intermittently is fundamentally flawed. The flaw is exposed when the employee tries to access a URL that is not recognised by the registered list. When this occurs one of two things can happen, either staff can have blanket access to unlisted sites or all unrecognised sites are blocked. If staff are given unfettered access to all sites on the Internet, it may effect that staff memberıs productivity and the company may be exposed to objectionable material and this may also have legal ramifications. In the event that all unrecognised URL's are blocked, employees may be denied access to web content that may be necessary to carry out their job and hinders productivity