Content-control software

A content filter is a program that is used to filter certain unwanted content (e.g. data, images, audio files, ...) before it can be transferred from one network to another.

That is why a content filter is usually located at the interface to the network input and controls the data coming into the network before it can reach the part of the network accessible to the users.

These content filters are often used by companies, governments and organizations to recognize certain content according to characteristics and properties. So that this recognized content can be blocked and excluded from the own network. The goal is to protect the users of a network from dangerous (e.g. terrorism propaganda) and illegal content (e.g. copyright infringements). It can also improve the productivity of users (e.g. less distraction) or increase the technical security of a network (e.g. against industrial espionage).

An example of the use of content filters by governments is the blocking of illegal, offensive or harmful websites.

Properties and functions

A filter that can be set in detail facilitates control over the content viewed. Partly unnoticed by the user, this system blocks or filters, for example, web pages or e-mails with regard to individual words, typical phrases, images or links. Content filters are part of the BSI recommendations for the operation of secure mail servers.

Usually, these areas are configured by means of categorization, using lists (databases), such as the rejecting blacklist or a whitelist that nevertheless allows access. Typical examples of predefined categories are pornography or right-wing extremism.

Simple content filters

Simple filters only check the occurrence of certain selection criteria. The disadvantage of using very simple filters is that every page with corresponding words is blocked. This can also block pages that use blocked words in a different context. For example, if the word "sex" is blocked in any form, not only will any page in general that uses this word - even in an innocuous context - be filtered, but possibly also the page for the village of Sexten or the Wikipedia entries right-wing extremism and random experiments. This is also called the Scunthorpe problem.

Another simple method is to block certain ports or network protocols.

Intelligent content filters

To circumvent disadvantages of simple methods, intelligent filters work with weightings and reject a page only when a certain relevance is exceeded. In addition, they also use heuristic methods and, in the case of e-mails, greylists.

In contrast to a URL filter with a high hit probability, an intelligent content filter is able to correctly assess web pages or e-mails. Likewise, URLs that are found to be good can be cached for a certain period of time to avoid the time-consuming process of checking for repeated requests.

Restrictions

Simple filters can only check URLs, texts and images classified according to certain standards. Texts that are displayed in image form can only be checked if the filter also has support for OCR text recognition. Of course, it is also only possible to a limited extent to keep "blacklists" of all undesirable pages.

Advanced systems also combine an antivirus program with the content filter in order to detect malware that has been introduced via scripts or images, for example.

Today (2014), content filters are often offered as a separate service for a fee; the service providers take over the updating of the filter lists and the intelligent filtering according to configurable criteria. Significant costs are incurred per year and Internet user.