Internet Filters and Censorship: distinguishing between oppression and age-appropriate information dissemination.

Filtering Methods

This section will help acquaint you with what filters are and how they work. There are a variety of methods employed by filtering companies to identify unacceptable Internet sites, and essentially these methods are what the various types of filtering software use to prevent or reduce access to unnaceptable materials.

�

• Blacklists / blocklists / host site blocking - lists created by filtering company staff who review sites and URLs that have been determined to be objectionable by keyword searches. Then those sites are blocked with the filter, but everything else will be allowed to pass. The lists are proprietary to the vendors, and will not be revealed to filter customers. This type of filtering, or list creating, is more of an art than anything else. While it is possible to have automated software tools do preliminary sweeps of sites and flag those with questionable content, a human still has to review them and make the final determinations. This leaves the door open for personal interpretation and bias. Two main difficulties that may arise with site selection are first: that sites with content that did not appear questionable to the robot software, yet definitely contain material that list makers wish to block, may slip by; and second, that sites may be blocked simply because a certain aspect of a website's makeup fits the site selector criteria for blocking. Local Site Lists can be used to override vendor settings.

�

• Image Recognition Software - software that attempts to recognize and block inappropriate images.

�

• Keywords in the body of the site, or metatags (identifying data appearing in the source code of a web page). Keyword method uses automated searches to locate objectionable words within the site text. Keyword filtering is dependant upon a lexicon of objectionable words and/or phrases. The vendor or (if allowed) the user determines which words are objectionable, and the words are primarily sexual in nature. Keyword blocking is also known by a variety of terms including "content analysis", "content identification", "phrase blocking", dynamic document review". Keyword blocking has a reputation of being less than satisfactory, as it is based on pattern-matching. The main problem is that the software functions on the basic idea that a single word never has more than one meaning.

�

• Keywords in context - algorithms and language analysis software make determinations regarding the use of an objectionable word in context. The purpose of this is to prevent blocking of legitimate word usage, such as breast cancer. The same issues presented in the keyword section above apply here, except that supposedly the pattern matching is more sophisticated when keyword searching in context is performed. But the success rates of this form of keyword searching are little more successful.

�

• Protocol Blocking - This type of filtering software can block e-mail, chat, newsgroups and FTP sites. This is because the format / Internet location (protocol) can be blocked.

�

• Rating Systems - A self-imposed Internet ratings system utilized to identify websites (similar to tv and movie industry ratings). The Platform for Internet Content Selection (PICS) was created to determine and devolop these ratings. Some companies that use these ratings include AOL, CompuServe, and certain filtering companies.

• Web Rating Systems (PICS, or Platforms for Internet Content Selection) - this technology enables rating systems based on metadata information (metadata = information about information). Ratings can be either created by ratings companies or web sites can be self-rated. There is no industry standard, and therefore rating systems can be somewhat arbitrary and contradictory. For an example of a web rating site see RSACI at http://www.rsac.org , where you can also rate your own site. What the ratings systems do is apply metadata labels for sites with such categories as Language, Violence, Sex, etc., with a variety of subcategories for each. The metadata labels can be embedded in the HTML code in their sites, where it remains unseen. Once the PICS metadata label has been added, then filters which have been enabled for PICS can source the code and determine whether a site fits the blocking criteria set for that particular filter. This allows the site to either be passed through the filter or blocked as is appropriate. One problem with PICS labeling is that labels can be kept secret, and again there is the issue of knowing what is being blocked. Another is that with self-rating site authors can include inappropriate labels for their sites. And for PICS to work completely it would be necessary for every site on the Web to have been rated. Considering the size and scope of the Internet, it is unlikely that this will happen in the near future.

�

• Whitelists - a reviewed list of Internet sites that have been approved by vendors. The filter in essence denies access to all sites and URLs but those on the whitelist

�

Filtering Problems

• Internet size and scope - The Internet keeps expanding at an exponential rate, making it difficult (if not impossible) for filtering to keep up with the ever-increasing size and scope of the Internet.
  • Search engines index less than a third of the Internet.
  • Approximately 25 new web pages are added to the Internet every second.
  • Approximately 7.3 million new web pages are added to the Internet every day.
  • The World Wide Web doubles in size approximately every eight months.

�

• Leaks - It is possible for objectionable sites or information to get past filters.

�

• Legal Issues - Despite the beliefs of some, pornographic and other objectionable information/materials are not illegal.

�

• Overblocking - Appropriate and non-objectionable sites are often blocked during the keyword filtering process.

�

• Practical Issues - the American Library Association has stated that "there is no proven technology that both blocks out all illegal content and allows access to all constitutionally protected material", and has compared filtering to a "technology arms race".

Legal Aspects for Public Libraries and Schools

Although there have not yet been any specific federal or state court rulings dealing with filtering software, there have been several cases suggesting information access must be protected. This list contains legislation affection censorship, and leading to the state in which censorship new stands. For a more complete list of historical censorship legislation see Censorship by Gail Blasser Riley (1998.)

• Obscene Publications Law of 1865 was first passed to address the issue of obscenity sent through U.S. mail. It was later replaced by the Comstock Act.

• The Freedom of Information Act of 1966 (FOIA) required that the public be allowed access to government records. Only nine areas were exempt, including an individual's right to privacy. The law provides sanctions for documents illegaly withheld by the government.

• The Privacy Act of 1974 went further, mandating the release of information to individuals on whom government agencies had kept files, in other words, people had a right to know what the government knew about them, and the government was forced to provide that information.

�

• The Communications Decency Act of 1996 was struck down in U.S. District Court for the Eastern district of Pennsylvania, stating in part that "governmental regulation of the content of speech is more likely to interfere with free exchange of ideas than to encourage it. The interest in encouraging freedom of expression in a democratic society outweighs any theoretical but unproven benefit of censorship." Often called "Black Thursday" by the Internet community,

�

• Louden County Public Library was forced to remove filters from its public Internet terminals in 1998 after they were declared illegal in U.S. District Court. This was in part due to the nature of the filter type being used by the library. The XStop Librarian II filtering software made use of site selection criteria that was kept secret, a key issue in the judge's decision declaring use by the library of the filter was illegal. Judge Brinkema stated "It has long been a matter of settled law that restricting what adults may read to a level appropriate for minors is a violation of the free speech guaranteed by the First Amendment and the Due Process Clause of the Fourteenth Amendment". The judge made it clear that limitation of adults to a speech level appropriate for children was illegal.

�

• A California ruling in March of 2001 pertaining to the Livermore Public Library made clear that libraries are not responsible for access by minors of pornographic materials on the Internet. The ruling stated "No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider."

What CIPA means for public schools and libraries is that either proof of actions showing movement towards compliance be shown, or loss of funding for communications will occur, in some cases a disastrous loss of approximately 90%, and in most cases 20% or more, depending on the community circumstances.

�

Bibliography

Aftab, Parry (2000). The Parent's Guide to Protecting Your Children in Cyberspace. McGraw-Hill, New York.

Bess: the Internal Retriever (N2H2). http://www.n2h2.com/

Children and the Internet: Guidelines for Developing Public Library Policy (1998). The American Library Trustee Association (ALTA), the Association for Library Service to Children (ALSC), and the Public Library Association (PLA).

Computer Professionals for Social Responsibility: Filtering FAQ. http://www.cpsr.org/filters/faq.html

Consumer Reports Online: Internet Filters. http://www.consumerreports.org/main/detail.jsp?CONTENT%3C%3Ecnt_id=18867&FOLDER%3C%3Efolder_id=18151&bmUID=993422068601

Cyber Patrol. http://www.cyberpatrol.com/

CYBERSitter. http://www.cybersitter.com/

FindLaw Online Legal Database http://www.findlaw.com/

Information Filtering of the Congressional Record: Explaining Variation in Adoption and Intensity of Use. http://www.ils.unc.edu/gants/dissertation.htm

Intellectual Freedom for Children: the censor is coming (2000). The Intellectual Freedom Committee, Association for Library Service to Children..

Internet Filters 101: Evaluating Internet Filters for the Public Library http://tln.lib.mi.us/~amutch/pro/filter/lom2001/

Lexis-Nexis Research System http://www.lexis.com/

Net Nanny. http://www.netnanny.com/home/home.asp

Net Shepherd. http://www.netshepherd.com/main.htm

Oder, Norman; Rogers, Michael (1999). "Suit Challenges 'Adult' Book Rule." Library Journal, 9/15/99, vol. 124 Issue 15.

Platform for Internet Content Selection (PICS): Internet site ratings. http://www.w3.org/PICS/

Parental Controls / Internet Filters. http://www.microweb.com/pepsite/Software/filters.html

Schneider, Karen G. (1997). A Practical Guide to Internet Filters. Neal-Schuman Publishers, Inc., New York

Senate Reports Online via GPO Access [wais.access.gpo.gov], 105th Congress 2d session, Senate report 105-226.

Smart Filter. http://www.securecomputing.com/index.cfm?skey=85

Smut Sieve: why Internet filters aren't enough to protect your kids. http://www.zdnet.com/anchordesk/stories/story/0,10738,2690741,00.html

Surf Control. http://www1.surfwatch.com/home.html

The Big IF. http://www.turnercom.com/if/index.html

TIFAP: The Internet Filter Assessment Project. http://www.bluehighways.com/tifap/

X Stop. http://www.xstop.com/

Pamela Force

Independent Study IRLS699

University of Arizona

School of Information Resources and Library Science

�

last update 4/8/02

created by Pam Force