 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
E-Rate and Filtering: a Review of the Children's Internet Protection Act.
Subcommittee on Telecommunications and the Internet
Mr. Chris Ophus
Introduction- My name is Christian Ophus, I am the co-founder and President of FamilyConnect, Inc. and S4F Technologies, Inc., a filtering technology provider founded in 1997 and headquartered in Tulsa, OK. In addition to my corporate duties, I currently serve as President of the Internet Safety Association, founded in September 2000 and headquartered in Washington D.C. The ISA (Internet Safety Association) was created by leaders in the Internet Content Management Industry to promote safe use of the Internet for all users. I would like to thank the U.S. House of Representatives Committee on Energy and Commerce, Subcommittee on Telecommunications and the Internet for inviting me to submit testimony. I will focus my comments specifically on filtering & Internet content management technology, offering background, current approaches and tools, and future developments.
Technology Protection Measures- Why are they necessary?The Internet is truly the most comprehensive and unique mass medium in the history of communication. The Internet is rapidly becoming the convergence of all other forms of communication. Television, radio, print, postal service and telephone service, are all available via the Internet. But even more amazing, is that the Internet has become the new backbone of these other communication mediums, ensuring that the Internet industry is here to stay. Our dependency upon this new medium has flourished, especially in the past decade. The Internet is an emerging technology that has it's own set of problems. The Internet is essentially an open network with a common language that allows anyone worldwide to access and transmit information. It is essentially a public forum, which fosters the free transmission of information and ideas. One of the sacred cornerstones of the founding fathers was to preserve the free transmission of ideas and information. That is why the very first amendment covered this issue. However, there are obvious exceptions to the first amendment. Information that is obscene, illegal and harmful to minors is not protected under the first amendment. Outside of the Internet, this type of information in any other medium is prosecutable under existing laws and regulations. To understand why illegal content via the Internet has become so controversial is puzzling. One might ask: What makes the Internet immune to existing laws and statutes that are already in place to protect individuals from material that is deemed detrimental in nature? Although the Internet is a viable tool for business, education and commerce, there is a significant amount of obscenity and illegal information. The goal is to limit access to this type of material without affecting the overall Internet experience for the user. Filtering technology is the best alternative to solving these issues. Historically, there has been controversy concerning the effectiveness of filters. The rapid growth and dynamic nature of the Internet make Internet filtering a constant moving target. In the mid-nineties, a few companies emerged in an effort to offer technological solutions to the ever-expanding problem of detrimental and illegal activity on the Internet. The first approach relied on artificial intelligence to block access to pornographic or objectionable material. These systems were based on keyword filters that would filter incoming data and look for words such as "sex", "XXX" or "breast". This type of approach was, in fact, good at identifying pornographic & illegal websites, but inadvertently blocked legitimate site searches such as "Middlesex", "Super bowl XXX" or "chicken breast recipes", etc. To solve this problem, new ways of filtering would have to be developed. Many opponents of filtering use the argument that filters still make these kinds of mistakes. Today's technology has risen far above these early products by using computers that scour the Internet coupled with human review to ensure a high level of accuracy. In fact, today's Technology protection measures are more advanced than ever before. Not every filtering product is the same. In the same way that there are different types of automobiles, some have more features than others, some are more expensive and then there are some that were created with specific purposes in mind. If your desire were to race in the Daytona 500, then you would not drive a Yugo. If your goal were fuel economy, you would not drive a Hummer. Similarly, there are different types of filters for different objectives. Some are less expensive and offer less protection and less control. At the same time, there are filtering products that have been specifically designed to operate in a more commercial application such as large corporations, schools and libraries. To ensure successful lasting implementation of a technology protection measure, you must fit the product with the application. Opponents of filtering have misled the public into believing that filtering does not work, or more accurately, does not work well. The justification for this claim has been a few isolated studies where the testing criterion is questionable and the results generalized. In March 2001, Consumer Reports published an article about filtering technology where 6 off-the-shelf filtering products were tested. The results indicated that the tested products did poorly when the testing criterion was applied. The article proceeded to question the government's imposition of filtering on schools and libraries through the Children's Internet Protection Act, citing that the test results were clearly negative. In response to the article, I wrote the editor of Consumer Reports on February 23, 2001 and questioned the products tested and the criterion used to test the effectiveness. Here is an excerpt of that letter: "First, the objectionable contentsite sample used, 86, was obviously but a small fractionin comparison to the vast number of adult and illegal websites on the web. To effectively test any filter, a more appropriate sample might have been 10,000 or even higher.
Second, a thoughtful set of criteria should be established in the selection of sites to be tested to ensure that the sites chosen are a statistically accurate representative sample of the range and type of objectionable sites found on the web. Your article did not indicate what criterion, if any, was used to determine which 86 sites were to be used. For example, we do not know if the author searched for 86 obscure sites or chose a random sample from a popular search engine. The answer to that question would dramatically affect the outcome of your informal survey.
Third, only six of the 141 filter-related products listed on the popular information website www.getnetwise.org were tested. The products tested, with the exception of AOL's parental controls, are client-side products. No server-side filter systems were tested. Also, some of the most popular filter programs were not included in the test.
Fourth,none of the filters tested are those typically used in the educational space. Filters such as N2H2, X-stop, I-gear, S4F and Web Sense were not even mentioned, and these products represent the vast majority of the access-control market share. Would it not be reasonable to test those products that are most commonly used and perhaps those who have made the greatest advancement in creating solutions that work for everyone?
Fifth, the test conducted did not include one of the most important aspects of filtering, the ability of the software to be overridden or bypassed by web-savvy kids. A filter can be a false sense of security to a parent or educator if it can be easily bypassed. Features such as this contribute greatly to the overall value and effectiveness of a filter.
I hope you can see how these seemingly innocent oversights lead to erroneous, generalizedconclusions. The fact is, there have been significant advancements by many companies even in the past year that validate the claim that filtering works and is effective in protecting children from illegal and dangerous information."
In response to my letter, I received a return letter dated March 7th, 2001 where the editor admitted that the products tested were from the consumer level and not those used in the educational space. "We are, however guilty of testing only so-called client-side software. Since our founding in 1936, we've focused on testing products available to consumers at the retail level. It is not part of our brief to test software sold exclusively to schools or libraries. By analogy, we would test garden hoses, sponges and auto polish, but not commercial car-wash equipment."
I encouraged the editor to consider a more comprehensive test where some of the more popular and broadly used filters could be included. I am sure the results would be entirely different.
David Burt, in his written testimony before the COPA commission in July of 2000, cited several larger studies of Internet filtering products where the outcome of filtering effectiveness was quite different.
In the Dangerous Access, 2000 edition by David Burt, the filter product, Bess, used at the public library in Cincinnati and Hamilton County wrongly blocked sites only .019% of the time.
A study by Michael Sims "Censored Access in Utah Public Schools, 1999" found error blocking rates at .036%. These numbers are a far cry from so-called tests being highlighted by filtering opponents.
Christopher Hunter, a COPA panelist said: "The majority of reports about Internet content filters being both under inclusive and over inclusive have come from journalists and anti-censorship groups who have used largely unscientific methods to arrive at the conclusion that filters are deeply flawed."
Current Approaches to Content Filtering
There are two typical approaches to filtering - inclusion filtering, and exclusion filtering. Inclusion Filtering- White Lists With inclusion filtering, Internet users are permitted access to particular "allowed" sites. This type of filtering can be 100% effective - assuming the person or organization that has compiled the white list that shares the same set of values as the Internet user. Because of the global nature of the Internet, it is difficult to create with a globally accepted set of criteria. The main drawback of inclusion filtering is that the "acceptable list" would have to be enormous to be accurate. The creation of a blocked list tends to be more manageable. Exclusion Filtering Exclusion filtering is based on black lists (or block lists) of objectionable sites. This is a more common form of filtering than inclusion filtering, and has the advantage that black lists will invariably be smaller than white lists. A second advantage is that unrated sites are presumed to be innocent till proven guilty, and so do not need to be automatically excluded. Both types of content filtering require a constant effort to maintain a valid and updated list for use by the user. The most effective approach is to use the benefit of computer technology, coupled with unique capabilities in human review.
What Content Can Be Blocked? In the early days, companies offered 1 or more categories of blocked sites, offering little or no control to the end-user. Today, most companies offer multiple categories and varying levels within these categories, giving complete control and flexibility of application to the end-user. Some filtering providers offer as many as 35 categories allowing the administrator complete local control over what is being blocked. Here is an example of a typical category listing from N2H2: Adults
Only, Auction, Chat, Drugs, Education, Electronic Commerce, Employment Search,
For Kids, Free Mail, Free Pages, Gambling, Games, Hate/Discrimination, History,
Illegal, Jokes, Lingerie, Medical, Message/Bulletin Boards, Moderated,
Murder/Suicide, News, Nudity, Personal Information, Personals, Pornography,
Recreation/Entertainment, School Cheating Info, Search, Search Terms, Sex,
Sports, Stocks, Swimsuits, Most of the above categories are not classified as illegal or detrimental in nature, but give the user a wide range of control when determining what information is appropriate for the viewer or more commonly, for what application the filter is being used. An employer may want to block access to job sites or other non-work related sites to reduce employee Internet abuse in the workplace. Several studies have indicated that loss of productivity from Internet use has cost employers billions of dollars each year. The point is filtering products today offer the user a wide range of options and combinations that allow the user to determine what is and is not blocked. In the educational space, the local school board can determine what information is appropriate to block based upon community standards, federal laws and the individual states harmful to minors statutes. How Filtering Is Accomplished There are several approaches to filtering content. As technology has progressed, the most effective methods have been improved, new ways to filter have been developed and many products have taken the best features from each approach and created a hybrid of several methods. There are four primary methods that are used in varying degrees. URL Filtering This is the most common, and most effective form of filtering, and involves the filtering of a site based on its URL (i.e. its address). It provides more fine-grained control than packet filtering, since a URL can specify a particular page within a large site, rather than specifying the IP address of the computer that hosts the Content. S4F Technologies adds an average of 5,000 - 7,000 new URL's to its database each week. Computer spiders scour the Internet using a sophisticated search mechanism that collects potential sites for human review. Spidering computers run programs that systematically read through the World Wide Web and collect URL's (Uniform Resource Locators) that match a particular set of criteria established a filtering department. These computer can run 24 hours a day and collect potential candidates to be added to the database. However, spiders are not perfect, and using spiders alone as the mechanism for fortifying a blocked site database would result in overblocking. That is why human review must be used when accurately building a blocked database. During the human review process, using custom browsers, sites can be positively identified and properly added to the database. As soon as a site is added, it is active in the blocked list for all to use. If a site is inadvertently blocked, it is reviewed and a decision is made within 24 at the most. If the site contains Child Pornography it is automatically forwarded to the National Center for Missing and Exploited Children. One of the challenges facing filtering departments is managing the constant change of the Internet. When a website is reviewed, it may not contain obscene material, but at some later point, the author of the website may change the content that now would be considered inappropriate. Conversely, a site with content that may have at one time been considered pornographic or illegal could change and be perfectly acceptable. So, in addition to keeping up new sites that come online daily, filtering departments must constantly review those sites that are already categorized. Considering the ongoing task of Internet content data management, coupled with the constant change in the Internet snapshot, filtering companies do an amazing job of keeping up. Keyword Filtering Keyword filtering was the first generation of filtering. With keyword filtering, content is scanned as it is being loaded into a user's computer for keywords, which are included in a black list. A site is blocked if it contains any of the words in the block list. The advantage of keyword filtering is that it adds very little computational overhead. The main disadvantage is that it checks text only, and cannot block objectionable pictures, plus, some products filters are indiscriminate, as the context is not taken into account. However, one of the advances of S4F Technologies, is the development of IKSSB (Intelligent Keyword Search String Block out) where the keyword component operates as a secondary line of defense to the primary specific URL block out database, and has the ability to decipher the difference between a website containing pornography, and one that has text which contains the word pornography. For example the IKSSB can differentiate between searches for "breast" and "chicken breast recipes" or another example, the difference between "sex" and "Middlesex, England". Both of these examples have been tirelessly used by opponents of filtering to claim that keyword filters can block useful sites. IKSSB Keyword Search String Examples Blocked Not Blocked Sex Middlesex, England Sexually Transmitted Diseases Sex Education Sextant
Breast Chicken Breast Breast Cancer
Although S4F uses this filter component as a secondary line of defense, it exhibits the technological adaptation of filtering companies to remedy earlier filter problems. Technical issues regarding filters have been overcome by most leading companies in the filtering space.
Packet Sniffing Content is delivered over the Internet in packets of information. Each packet has the IP address of where it is going to, as well as the IP address of where it has come from. Packet sniffing involves examining the IP address of where the Content has originated. This approach moves the point of filtering to the level of the router offering increased speed and efficiency. There are several companies that are developing packet-sniffer products at this time. Image recognition filtering A handful of companies have produced filtering products that examine images as they are delivered to a user. This is a relatively recent approach, and relies on techniques such as the detection of skin tones, or indeed on the analysis of images themselves. It is computationally quite intensive, and computers will invariably experience difficulty in distinguishing between art and pornography. A photograph that is artistic in nature cannot be distinguished from that of obscenity. These types of value judgments can only be made by human review. Video and other streaming media further complicate the filtering task by supplying a constant flow of images to be examined for undesirable content. Where Does Content Filtering Occur? There are four technical components of filtering systems: browser-based, client-side software, proxy servers, and server-side filtering servers.
The security loopholes with client-side software are a concern. Many smart children can disable filtering software faster than a parent or teacher can install it. In addition, there are quick and easy programs written to disable the major companies' software with the click of a mouse. These programs are circulated among children who simply download it from the Internet, place it on a floppy disk, and pass it around.
Filtering functionality can be removed from the end-user's computer and placed on a server somewhere else on the Internet, called a proxy. With a proxy server, all website traffic must go from the end-user's computer through the proxy server, then to the rest of the Internet. Proxy servers offer more security than client-side software. All users must go through this proxy server to be able to access the Internet 'proper'. To do so, the client is required to configure their software to 'point to' this proxy server to be able to access Web pages and ftp files. a range of Internet-based Failure to do so will result in blocked access to the Internet. A proxy filter can be selective about what it blocks, and can be configured to block or permit access to services.
Browser Settings
Microsoft Internet Explorer provides content security settings for the Internet Content Ratings Association's RSACi ratings, the most popular ratings system on the market. However, if a site is not rated, it is not accessible. Popular sites that are not rated include ESPN, CNN, eBay, Amazon, and AOL. In fact, most sites are not rated, making them inaccessible to the user.
Some filtering software "decides" what to block based on how a site is rated -- not entirely unlike the way parents use movie ratings. This method offers fewer features and less precision compared to some of the higher-end server-side products. Hybrid Filters There is a new filtering method that utilizes the best features from each of the other methods. This hybrid system has varying forms. S4F Technologies patent-pending system uses a server-side component that works in tandem with a thin client-side software interface. By using more than one method, the user is able to take advantage of the benefits of server-side filtering, including real-time access to the most up-to-date database, the speed benefit and user-control features of client-side technology.
Future Advancements In Filtering Technology Filtering technology providers have dedicated thousands of man-hours and millions of dollars in research and development to create real solutions for schools, libraries, homes and businesses. At best, the filtering industry is only 7 years old. The advancements in technology over the past 2-3 years alone have brought about products that combine artificial intelligence, advanced algorithms, intelligent keyword databases, computer spidering technology, millions of websites accurately categorized. All of this, while increasing speed, efficiency and manageability through cutting edge system design and engineering.
Internet filtering is not foolproof. The dynamic of the Internet as it relates to filtering can be likened to virus detection software. Products in the virus detection industry use similar algorithms, they monitor packets being transmitted over networks, and they have extensive databases of known viruses and their signatures, yet these virus detection tools are not fool-proof, still network administrators world-wide use these programs to protect their networks because that can offer a high level of protection, even if it is not 100%.
It seems that the opponents of filtering technology wish to cast down the use of any filtering software because it might only be 95-99% effective. Opponents are trying to hold filtering software to a higher standard than other types of similar and related products. Windows and Macintosh operating systems, Internet dial-up connections, computer manufacturers and virtually any software application manufacturer all create and sell products that are not fool-proof and error free. That is why software companies continue to release updates and create new versions, to keep up with the ever-changing marketplace. It is an acceptable part of the computer industry.
Future filtering technology advancements will see the convergence of several of the approaches reviewed. Are There Other Alternatives To Filtering? Some of the opponents of the Children's Internet Protection Act have suggested that filtering is not necessary; rather, a strong education program that trains children how to have a positive Internet experience is all that is needed. Although I feel that education is a great way to teach children about the dangers of the Internet, it is surely no replacement for technology protection measures. The biggest problem is that much of the pornographic and illegal exposure to minors is accidental. The National Center For Missing and Exploited Children released a study where 1 in 4 minors reported viewing of unwanted material. It is a well-known fact that in an effort to increase viewer ship, operators of obscenity websites will use unrelated keywords and misleading URL's to attract unsuspecting users to their site. Once the image is viewed, the damage is done and the law has been broken. All the education in the world cannot stop that from happening. To illustrate this erroneous argument, consider drivers education. Millions of teenagers and adults each year take some form of driver's education or training. Yet the government has put seatbelt laws in place to protect people from harm. All the driver's education in the world cannot stop accidents from happening. Seat belt laws do not guarantee to protect the passenger 100% of the time, in the same way that Internet filters cannot ever guarantee 100% perfect performance, yet they are a great tool to divert the vast majority of Internet abuse in schools and libraries. Monitoring has been considered as an alternative to filtering. This approach places the burden of policing the Internet on educators and librarians who cannot possibly mange the activities of every Internet user. Once again, if sites are accidentally seen, the damage has been done. Can Existing Technology Protection Measures Meet The Requirements Of The Children's Internet Protection Act? The answer is a resounding yes. The Children's Internet Protection Act requires that a school or library select a technology protection measure, which they choose, not the government through a public hearing and the creation of an Internet safety policy. The local board determines what to block based upon Federal and state laws as well as local community standards.
This law encourages public education and empowers consumers and local authorities to work together to create a solution that is right for everyone. Schools and libraries have the affirmative duty to protect minors while in their custody. Using technology protection measures shows that educators are taking reasonable steps to protect their kids. Effective filtering technology exists and is effective.
The leading filtering products in the educational space already have the necessary functionality to meet the requirements of the law. Here is a profile of those products: CIPA related features comparison of the most popular filters in public schools and public libraries. (provided by David Burt of N2H2)
Evidence of Librarian Satisfaction with FiltersStatistics show a dramatic increase in filter use in libraries. A new study [1] by the U.S. National Commission on Libraries and Information Science shows a dramatic increase in the number of Public Libraries using Internet filters. In 1998, just 1,679 public libraries offering public Internet access filtered some or all Internet access. [2] In 2000, that number more than doubled to 3,711, [3] representing an increase of 121%.
One in four Public Libraries
offering public Internet access now use filters.
The most dramatic gains came in Libraries filtering some Internet access, which increased from 801 or 7.0% in 1998 [8], to 2,265 or 15.0% in 2000 [9]. Data from this study indicate that there has been a 65% increase in Public Libraries filtering all public Internet access since 1998. The number of Libraries that filter all access has climbed from 878 or 7.6% in 1998 [10], to 1,446 or 9.6%. [11] Nearly 1,500 public libraries (one out of every ten) filter all access today.
A Survey shows that librarians
and teachers are highly satisfied with filters.
An astonishing 90% of public librarians who used filters responded that "the software serves its purpose" either "very well" or "somewhat well". [13] The study asked both school and public librarians who used filters to rate their level of satisfaction with filtering software in several ways.
[14]
[15]
News stories and public statements made publicly by librarians and library patrons reinforce the research Claudia Sumler, Director of the Camden County (NJ) Library System:
She said that if a patron complains that the technology is blocking a legitimate site, librarians easily can override the controls. "If there are complaints, librarians are to deal with them right away," Sumler said. "We don't want to deny access.". "We think this works for us," she said. [16]
David C. Ruff, executive director of the Rolling Meadows (IL) Library: Expanding the filtering technology to block obscenity and pornography on the library's 20 public computers was based on the library's satisfaction with the Cyber Patrol software and the desire to simplify some administrative procedures, said David C. Ruff, the library's executive director.In the week since the filtering policy was expanded, patrons have not noticed the difference, Ruff said. [17]
Joan Adams, director of the Jefferson Parish (LA) Public Libraries: After several months of wrangling with software companies, parish officials on Thursday finally finished installing filtering software on about 100 computers, cementing the Parish Council's promise to do what it legally can to keep perverts and smut out of public libraries. But for most computer users who sat quietly pecking away at their keyboards, the added restrictions were hardly detectable."I've gotten a lot of 'what if?' questions from the librarians," [Library Director Joan] Adams said. "But the average computer user doesn't even notice it is there."
So far, the WebSense software does not seem to be slowing down the speed of library computers, a common side effect to installing filtering software, library network administrator Dwight Bluford said. The software program also seems to be fairly on target with the sites it blocks. That's because WebSense searches the content of Internet Web sites to determine if there is offensive content, not the keywords, he said. "It seems to be working well," Bluford said. And because it can be locally manipulated, "we also have the ability to immediately block a site if we get a complaint from a patron, or to unblock a legitimate site if it is blocked." [18]
Library patrons and staff at the Plano (TX) Public Library:
Erin Noll Halovanic, Information Systems Librarian at Kenton County (KY) Library: Halovanic says if a customer complains about not being able to access a site that's supposedly suitable, she reviews it on an unfiltered staff PC and unblocks the site if she finds it appropriate for the library. And that seems to be a good compromise for Halovanic who admits, "As a librarian, filtering absolutely curdles my blood. It goes against my training as a librarian and my belief in librarianship. However, when it comes to the choice between pandering sexual materials and between protecting people's personal rights, I choose filtering over the alternative." [20]
Margaret Barnes, Director, Dallas (OR) Public Library: After much conversation and serious reflection, it was determined that a workable approach, enabling the Dallas Library to furnish access to the public, would be the installation of a filter system on all public Internet stations...During the almost 1 1/2 years that we have been providing this service we have had no one formally or really informally register an objection about a filter system being in place on the workstations. We have received countless positive comments about this service from all ages in our community. [21]
Judith Drescher, Director Memphis-Shelby County (TN)Library: The library system's switch to pornography-blocking software has gone so smoothly that it could be considered a nonevent. The Memphis area's chief librarian, Judith Drescher, told a Shelby County Commission committee Wednesday that more than half the 26 public queries about blocking software had nothing to do with the new software.In a report given to the commission's education and libraries committee, Drescher stated, "Since installation, the library has received no requests from the public to review and block a site. Library staff has submitted five sites for review, all of which were blocked." [22] FOOTNOTES FOR PRODUCT SURVEY 1. N2H2 offers six sex-related categories: "Adults only", "Lingerie", "Nudity", "Porn", "Sex", and "Swimsuits". Additionally, N2H2 has four "Allow exception categories" related to sexual material: "Education", for sexually explicit material that is of an educational nature, "History", for material of historic value, such as the Starr Report, "Medical", for material such as photographs of breast reduction surgery, and "Text", for pornographic or sexual material that only contains text. Category descriptions available at http://www.n2h2.com/solutions/filtering.html
2. WebSense offers five sex-related categories: "Adult content", "Nudity", "Sex", "Sex Education", and "Lingerie and Swimsuit." Category descriptions available at http://www.websense.com/products/about/database/index.cfm
3. Cyber Patrol offers five sex-related
categories: "Partial Nudity", "Full Nudity", "Sexual
Acts", "Sex Education." 4. I-Gear offers six sex-related categories:
"Sex/Acts", "Sex/Attire", "Sex/Personals",
"Sex/Nudity", "SexEd/Advanced", "SexEd/Sexuality"
5. Smartfilter offers three sex-related categories: "sex", "nudity", "obscene", "mature" Category descriptions available at http://www.securecomputing.com/index.cfm?sKey=86
6. X-Stop offers three sex-related categories: "R-rated", "obscene", "pornography" Category descriptions available at http://www.8e6technologies.com/docs/Manual_nt_proxy45.pdf
7.
N2H2 end users who feel they are unfairly blocked can request
a review, or request a site be blocked at http://www.n2h2.com/solutions/request_review.html 8.
WebSense end users who feel they are unfairly blocked can
request a review, or request a site be blocked at http://database.netpart.com/site_lookup/
. Users may also test a site to see if
it is blocked or not. 9.
Cyber Patrol end users who feel they are unfairly blocked can
request a review, or request a site be blocked at http://www.cyberpatrol.com/cybernot/
Users may also test a site to see if it is blocked or not. 10.
I-gear end users who
feel they are unfairly blocked can request a review, if the system
administrator has created a custom block page.
This process is described here. 11.
Smart Filter end users who feel they are unfairly blocked can
request a review, or request a site be blocked at http://www.securecomputing.com/index.cfm?sKey=234
Users may also test a site to see if it is blocked or not. 12.
X-Stop end users who feel they are unfairly blocked can
request a review, or request a site be blocked at http://www.8e6technologies.com/submit/index.html 13.
"Worldwide Market for Internet Access Control",
Chris Chistensen, IDC, 2000. Page 11. 14. "School Library Journal's School Internet Filtering Survey", Dr. Ken Haycock, Cahners Research, August 2000. Page 19.
FOOTNOTES FOR SURVEY AND QUOTATIONS
1. PUBLIC LIBRARIES AND THE INTERNET 2000: SUMMARY FINDINGS AND DATA TABLES. A report based on research sponsored by the U.S. National Commission on Libraries and Information Science (NCLIS) and conducted by John Carlo Bertot and Charles R. McClure. NCLIS Web Release Version, September 7, 2000 (visited February 8, 2000) < http://www.nclis.gov/statsurv/2000plo.pdf > (hereinafter "INTERNET 2000"). 2. 2. U.S. NATIONAL COMMISSION ON LIBRARIES AND INFORMATION SCIENCE, MOVING TOWARD EFFECTIVE PUBLIC INTERNET ACCESS: THE 1998 NATIONAL SURVEY OF PUBLIC LIBRARY INTERNET CONNECTIVITY. A report based on research sponsored by the U.S. National Commission on Libraries and Information Science and the American Library Association and conducted by John Carlo Bertot and Charles R. McClure. Washington, DC: U.S. Government Printing Office, 1999 (visited February 8, 2000) < http://www.nclis.gov/statsurv/1998plo.pdf > (hereinafter "THE 1998 SURVEY"). Out of a total population of 11,519 public libraries providing public Internet access (see Figure 8, p. D-10), 878 or 7.6% filtered all terminals (see Figure 48, p. D-50), and 801 or 7.0% filtered some (see Figure 49, p. D-51). 3. INTERNET 2000, at Figure 11, p. 18. Out of a total population of 15,128 public libraries providing public Internet access (see Figure 4, p. 11), 1,446 or 9.6% filtered all terminals (see Figure 11, p. 18), and 2,265 or 15% filtered some (see Figure 11, p. 18). 4. INTERNET 2000, at Figure 11, p. 18. 5. THE 1998 SURVEY, at Figure 48, p. D-50, and Figure 49, at p. D-51. 6. THE 1998 SURVEY, at Figure 8, D-10. 7. INTERNET 2000, at Figure 4, p. 11. 8. THE 1998 SURVEY, at Figure 49, D-50. 9. INTERNET 2000, at Figure 11, p. 18. 10. THE 1998 SURVEY, at Figure 48, D-50 11. Internet 2000, at Figure 11, p. 18. 12. SCHOOL LIBRARY JOURNAL'S SCHOOL INTERNET FILTERING SURVEY by Cahners Research, conducted by Dr. Ken Haycock of the University of British Columbia. August, 2000. (hereinafter SCHOOL INTERNET FILTERING SURVEY") 13. SCHOOL INTERNET FILTERING SURVEY, at Table 16, p. 9. 14. SCHOOL INTERNET FILTERING SURVEY, at Table 15, p. 8. 15. SCHOOL INTERNET FILTERING SURVEY, at Table 16, p. 9. 16. "Philadelphia-Area Library Found Internet Filters Far from Simple", The Philadelphia Inquirer, March 8, 2001. 17. "Meadows library expands filters on Internet access", Chicago Daily Herald February 25, 2001. 18. "Library's new Internet filters in place; Program installed; few seem to notice", The Times-Picayune (New Orleans), January 30, 2001. 19. "Internet filtering accepted; Libraries quietly implement policy", The Dallas Morning News, June 30, 2000. 20. "I-Gear for Education Success Stories: Kenton County Public Library", Symantec Website, available at http://www.symantec.com/sabu/igear/igear_educ/story_2.html 21. "Surfwatching the Internet", by Margaret Barnes, Oregon Library Association Quarterly, Volume 3, Number 4 - Winter 1998. 22. "Porn-Blocking Software Works at Library", The Commercial Appeal (Memphis), January 6, 2000.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
This
type of method is typically marketed at the consumer level. Filtering can be implemented by placing a
software program on the end-user's computer. The software then runs while the
user is online, performing the particular filtering functions. Client-side
software may require the end-user to configure the software and download
updated website lists. 
Proxy Servers
Filters
using built-in browser settings typically uses a ratings system. These systems are less intrusive but
typically less accurate. 
