Schools Need Better Web Site-Filtering Practices

MATTHEW WOLPA

March 19, 2006

The founders of the company N2H2 probably would not have named their Internet-filtering software after their family dog, Bess, if they had known that students across the U.S. would come to hate the name and the friendly logo portraying their pet.

Bess is a popular brand of censorware, the name given to filtering software used by more than 40 percent of U.S. public schools and libraries. Its original maker, N2H2, was bought by the Secure Computing Corp. in October 2003.

Bess is installed on a server connected to all of the computers in the Stayton School District. When a user tells a Web browser to display a Web site, Bess checks the requested site against a list of sites that aren't allowed.

If the site isn't on the list, it can be viewed. If it is, the user is taken to a block page with the Bess logo that states which offending categories the requested site falls into. Some of the more than 30 categories include pornography, profanity, visual search engine, games and recreation/entertainment.

The version of Bess used by my school has many flaws, the most important of which is the large number of false positives, or sites that are blocked when they shouldn't be.

Some sites are blocked for strange or false reasons, leaving students scratching their heads as to why something on the Massachusetts Institute of Technology Web site is blocked for "jokes." Harmless translation sites also have been blocked, as well as anti-censorware sites that criticize Bess.

Although it may seem as if Bess is too strict in its blocking, the software also lets a large amount of inappropriate content through. The essential weakness of the software is that it relies on a blacklist of "inappropriate" sites and is unable to determine by itself if a Web site is allowable or not.

Widely publicized gaps in the blacklist have been plugged, but Bess still is vulnerable to new sites that haven't been added to the blacklist yet or sites that Secure Computing isn't aware of.

Bess also confuses students by failing to give the real reason for which some sites are blocked. Google's new video search feature, for example, is blocked because it's a "visual search engine," implying that it could return image results from blocked sites.

Google Video, however, is a database of videos stored and filtered by Google. The site clearly states that the videos have no pornographic or otherwise inappropriate content. Bess really blocks Google Video because videos are bandwidth hogs, slowing down all other Internet traffic on the same connection.

The best way to fix the mess that is the Bess censorware is to completely replace it. Start by giving each student a personal logon and password. Then use new software that would start out with a basic list of allowed sites such as Google, online encyclopedia Wikipedia and educational resources such as OSLIS, the Oregon School Library Information System. It also would have a blacklist of proven bad sites.

If a student were to try to access a site unknown to the software, a screen would be produced that would inform the student that the visit would be recorded and the site would be reviewed by the school district's network administrator.

The student could then continue to the site as if there was no filtering software. If the site was discovered to be inappropriate by the district, it would be permanently blocked. If it was determined to be an appropriate, relevant or useful site, then it would be allowed permanently. Lists of good and bad sites could be built up quickly, and students who abused the system could be punished.

If school districts finally implement a better filtering solution, students will be able to use online resources more effectively.

Matthew Wolpa, 16, is a junior at Stayton High School. Contact him through education reporter Sarah Evans at sevans @StatesmanJournal.com or (503) 399-6856.