Several times a year, both staff and parents inquire as to why certain websites are blocked while others are not. These questions often lead to more general questions regarding how our web filter works and how reliable it is. What follows is a brief overview of web-filtering methodologies along with some specifics as it applies to Portage.
It has been estimated that the World Wide Web has over a trillion individual pages (no one knows the exact amount for sure). It is also estimated that more than 150,000 web pages are added each day. Given the ever changing content of existing pages and the constant addition of new ones, maintaining an accurate catalog of pages and their content is simply impossible. Any filtering strategy must acknowledge this fact and adapt as best it can. The three strategies most commonly used are content screening, whitelisting and blacklisting.
In this strategy, web-content is screened in real time as pages are requested. In its most basic implementation, should certain keywords or other specific patterns of data appear within the page, the filter prevents the page from loading. This is the least reliable method. For example, the filter might be set to block pages that mention the word ‘breast’. While this might block some pornographic web pages, it would also block access to medical information and poultry-based recipes. This strategy is also mostly ineffective in evaluating multi-media content. On the whole, this strategy is rarely used on its own.
In a whitelisting strategy, a list of appropriate websites is maintained by the filter, and only pages appearing on the list are accessible on the network. This is the most foolproof method, as access to any site must be preapproved by the filter’s administrator - although one should remember it is not completely foolproof as nothing prevents the owner of a pre-approved site from adding inappropriate content to said site. In practice, whitelists are impractical to maintain - there is far too much content on the web for a central source to screen and approve. As with content-based filters, whitelisting also tends to block access to many useful sources of information. As such, whitelisting also isn't used very often. Some applications include workplace computers intended for use with only a few sites (kiosks, point of sale terminals, data entry terminals) and filtering for personal devices assigned to younger learners. Whitelisting is also used for some multimedia applications.
In a blacklisting strategy, the filter allows access to any site except for those specifically listed on a ‘do not allow’ list. This strategy offers the best balance of access versus filtering out inappropriate content and as such is the most commonly deployed filtering strategy in work and school environments. Should an inappropriate site not appear on the blacklist, the filter’s administrator can quickly append it to the list, making the filter agile in the ever-changing internet environment.
In Portage, we use a mix of all three strategies to various degrees (mostly blacklisting). We subscribe to a content screening service that provides the bulk of our listing of sites to block. From there, based on feedback from our staff, we have customized the list to add some sites back in and blacklist several others. We have different tiers of filtering for staff and students. We also will exclude some sites to protect our internet bandwidth for instructional use. We routinely inspect our access logs to ensure that the filter is working the way it should, finding approximately two or three sites a year that need to be manually added to our list.
In summary, no system is foolproof. But, we are confident in our filtering implementation and the safety and security it provides.