Use of Content Filters on Internet in High School

¶ … Internet has grown exponentially since its first introduction to the public. The precursor to the Internet was the ARPANET. The Advanced Research Projects Agency (ARPA) of the Department of Defense (Carlitz and Zinga, 1997) and the National Science Foundation (NSF) were the primary creators of the ARPANET. Subsequently however, efforts from private entities and universities have helped develop the network infrastructure, as it exists today. "The goals of ARPA's 'Resource Sharing Computer Network' project were to develop the technology for and demonstrate the feasibility of a computer network while improving communication and collaboration between research centers with grants from ARPA's Information Processing Techniques Office (IPTO)." (Press, 1996) J.C.R. Licklider of MIT undertook groundbreaking work in developing computer interactivity. Later, he implemented his vision though time-sharing systems-affordable interactive computing. The effort of the NSF also helped to distribute the features of this new networking capability to all major universities and research centers in the U.S.A. (Carlitz and Zinga, 1997)

Purpose of this study

Changes in computer technology and the adaptation of computers to traditional educational needs have required that schools, universities and libraries make the necessary move to introduce the technology in their classrooms. As beneficial as the Internet has become in dissemination of information and knowledge, it is also ridden with websites that contain pornographic, violent and, often, disturbing information that can be objectionable to children and young adults. Access to state-of-the-art computers at affordable rates and Internet connections such as broadband (cable and DSL) and through telephone lines has brought the Internet to homes as well. The increased use of online options for dissemination of information by schools and the use of teaching mediums such as Blackboard have coerced the hitherto reluctant student to use the Internet. Library resources and research facilities are also increasing the use of the Internet by younger individuals all the time.

Teachers and schools have used the Internet introduce students to new ideas and knowledge, expeditiously. The drawback however, has been that the Internet has also become a convenient and restriction free medium for the nefarious, unethical and moral element of society to use as a medium for conducting their own business. Every element of the free and unrestricted nature of the Internet that has made it convenient and valuable for the use of the general public has also made it beyond the control of any single controlling agency. The Internet also transcends geographical boundaries. The ability of a single government or society to put restrictions on the content and the information that is to be displayed can be impossible.

This study identifies the need for content filtering in schools, colleges and universities. There have been varying efforts at content filtering and website blocking. These have been used to ensure that the student is offered reasonable buffering when dealing with the wide world of the Internet. There is no doubt that the environment within which the university operates has a great influence on the quality, range and the scope of information that is conveyed over its networks. There is also growing concern that blocking is required to ensure that the network being used is not bombarded with spam mail that can slow down the service that is provided. Another major concern faced by many democratic nations around the world is the scope and extent of filtration that can be legally mandated to be used in schools and colleges and the level of filtration that can be used in public libraries and government funded projects. In educational environments -- both formal and informal -- the role of the educator or the teacher is critical. The level of interaction that the teacher offers to help system administrators realize the extent and the range of blocking is also essential.

Research questions

This study seeks to identify the following information based on peer reviewed literature and governmental websites and forums that consider and review such issues. Using content filtering in high schools is essential. There are however, limitations and concerns with using these filters without adequately understanding the limitations and the flaws or drawbacks of this system. As the number of schools connecting and supporting Internet facilities on their premises is increasing, school officials often face the challenge of determining the level of control that is needed and the manner in which these controls are implemented. With this in mind, the following questions are considered significant;

1. What are the different types of filters and how can blocking be successful with the use of different filters?

2. How do Content Filters work?

3. What are the limitations of filters and issues of over-blocking and under-blocking?

4. How do filters impact the first amendment rights of the individual and the role that parents, society and schools play in determining the level of censorship that is needed?

5. Future of filtering and blocking of websites and the level of standardization and knowledge that can be built to support a unified system that can be used by schools and universities with minimum cost on resources and time

Limitations of this study

This study is based on peer reviewed journal and scholarly work only. It is based on secondary research and no primary data was collected and analyzed for the purpose of this study. This study is a generic study and does not identify any single school district or society requirements. In addition, this study considers school in all regions and does not make any single distinction between the level or grade of the school or the cadre of the students that might be using the system or their computer capabilities and knowledge of filtering systems that might be used. In addition, the data will not be restricted to any one country and information will be gathered from as many sources as possible within the scope and limitation of this study.

This study addresses some of the computational and the theoretical limitations of filter and blocking concepts. It does not attempt however, to describe or explain the code writing or the computational methods that can be used for software development or upgrades. The maturity levels and the users within the school system will also determine the type and level of filtering in place. While some aspects of the cost of maintaining a filtering system and the skill and capabilities of the systems administrator are discussed within the scope of this study, the true cost of maintaining this system is difficult to predict and determine. Only after a school or a school district implements such a program, and based on the level of maintenance required, will a true cost of the filtering process be determined.

Computers and technology are also more readily accepted by the younger generation. And as such, students might have more knowledge and insight into the operation of the filters and the blocking. This knowledge might enable them to circumvent the filters and obtain information through other sources. This paper does not identify these factors nor does it discuss the options that schools might have in preventing the problem of student sabotaging the filtering system or using other ways to access the blocked sites.

Conclusion

The technology revolution spearheaded by the introduction of the Internet and the availability of affordable personal computers has changed the manner in which we work, study and entertain ourselves. The evolution of connectivity to the Internet has undergone many stages. From the dial-up modem to the broadband technology as is used today, the connection speed available to the average Internet user is on the rise. Easy access to the Internet and the information that it makes available to the user is seen to be critical for the next age of computer use by citizens of any country. It has been increasingly observed that the computer industry, and the telecommunication and technology industry are converging. This convergence then presents opportunities and threats to users of the Internet.

Exercising complete control and monitoring the Internet is not possible: as it is a true open and global entity that was designed to allow information to be generated at any point and be accessed at any point. There are however many options that are available to users who wish to control the content and the availability of information. Countries like China, Iran and Saudi Arabia have very strict filtering and blocking rules. Access to information is limited and sites that are deemed 'inappropriate' (religion, sites providing information specifically to and about women, sites with sexual content, pages perceived to be hostile to Saudi Arabia and the policies of the region and some entertainment sites) are generally blocked. (Zittrain and Edelman, 2002) The country therefore, needs individuals from the region who are able to identify the topics that are of major concern and block the appropriate sites. The task of blocking also requires that a site with information that is inappropriate does not 'filter' in from other sources.

In the case of this study, when blocking of sites in high schools is referred to it is only applicable to blocking websites with pornographic and violent content. For an educational institution, the ability to connect to various online journals and libraries as well as access resources of other universities can help improve and growth the curriculum and the courses supported by school is critical and all effort should be made to ensure that over-blocking does not happen.

Literature survey

There is no doubt that the Internet has impacted the way information is collected, disseminated and made available free of cost (mostly) to all users. As useful as the Internet has been with all the information that it provides, some of the content is not desirable for all sections of society that use the Internet. Research indicates that parents, schools and society are increasingly looking for ways by which they can restrict the content of the Internet at the end-user point to ensure that the content is acceptable to the audience that is viewing it. (Delicato, Pirmez and Carmo, 2001) This task is becoming increasingly more complicated as the numbers of websites that are constantly launched are increasing exponentially. Filtering is also a complicated process. Software necessary for the filtering process needs to be able to separate different types of content using similar words more effectively. As the issues of offering safety protection against the unwanted websites that are out there is growing, the blocking or filtering that is employed is moving from "internal content filtering to network-level blocking techniques." (O'Donnell, 2005)

Blocking or filtering software is a mechanism used to:

"Restrict access to Internet content, based on an internal database of the product, or;

Restrict access to Internet content through a database maintained external to the product itself, or;

Restrict access to Internet content to certain ratings assigned to those sites by a third party, or;

Restrict access to Internet content by scanning content, based on a keyword, phrase or text string or;

Restrict access to Internet content based on the source of the information." (Rosenberg, 2001)

The objectives of blocking are not universal. Blocking or filtering might be undertaken for different reasons. This depends on a number of internal and external variables. The three main factors identified are;

1. "The supervisor: parenting styles differ, as do philosophies of management and government.

2. The recipient: what's appropriate for one fifteen-year-old may not be for an eight-year-old, or even all fifteen-year-olds.

3. The context: a game or chat room that is appropriate to access at home may be inappropriate at work or school." (Resnick and Miller, 1996)

The Platform for Internet Content Selection (PICS) ascertains the Internet conventions for label formats and distribution methods. The PICS introduces a difference between labeling and filtering. A label in this case offers a description of the content of a website. A filter makes the content inaccessible to some audience but allowing supervisors or parents to use the label as a predetermining factor. It does not however, mandate who should pay attention to these requirements. Syntax for rating the service is however provided. Users can identify the level of protection that is offered based on the information that is presented. Syntax for labels is also provided so that the computer can read them. A label may be digitally signed and may include a cryptographic hash of the associated document. An extension protocol is also provided so that the customer can transmit the label with the document. The "PICS specifies only those technical issues that affect interoperability. It does not specify how selection software or rating services work, just how they work together." (Resnick and Miller, 1996) Then, there is the question of the level of censorship and how it impacts the use of PICS ratings. As the definition of censorship is also often debatable, the extent of filtering that is sought is determined by the final objective that is to be achieved. For example -- children should not be able to access sites that contain nudity or inappropriate language.

According to Resnick, governments should be also responsible for offering some level of filtering that citizens can use for their own needs. (Resnick, 1999)

The government of a region could impose a labeling vocabulary and insist that all publishers use them accurately and effectively. While the government can set these rules, policing publishers can be challenging in lieu of balancing the first amendment rights of the individual. As such, labeling should only serve as a medium that users that review to ascertain if they wish to review the website or not. This concept is similar to the rating of movies that is offered to allow viewers to determine if it is appropriate for them. "A government could try to impose filtering criteria in several ways, including government-operated proxy servers (a national intranet), mandatory filtering by service providers or public institutions (e.g., schools and libraries), or liability for possession of materials that have been labeled a particular way." (Resnick, 1999) It is however, difficult to implement such an operation owing to the scale of efforts and the time and resources required to maintain such a system.

The U.S. Senate passed the Children's Internet Protection Act (CIPA) in December of 2000 in an effort to ensure that schools and colleges have the legal right to restrict certain websites and content from children and student bodies. (Cantos et al., 2001) Installing filtering software main aim is to "to protect children from obscene, violent, or otherwise inappropriate content on the Internet." It is important to recognize that the filtering software is not always foolproof. Many sites have been able to "work around" the filtering software. Critiques of this bill also are of the opinion that while the government has passed a bill to restrict Internet content in schools and colleges, it has not offered schools and universities with the required resources to do so. Recent surveys indicate that in the U.S. 95% of schools and 43% of public libraries use some form of filtering software for the Internet. (Resnick, Hansen and Richardson, 2004)

Parents are also increasing seeking out Internet service providers (ISP) that offer some level of control over the content of the Internet sites that is available. On the top of the list for filtering, are pornographic sites, violent and other offensive content. A pornographic website for the purpose of this study is one which contains sexually oriented contents -- verbal, pictorial or streaming video files.

Site blocking requires that all websites pass through an "Internet control point such as a firewall, proxy server, or caching device that immediately inspects each request to determine whether it should be allowed or denied." (SSS, 2002) In the actually working of such a system, all sites are logged or archived into a database. Alternatively, these sites are then matched up or parsed against a master database of websites that have been classified. If any blocking conditions are implemented, the questionable websites are then not allowed to load on the machine and are held at this inspection stage. It is possible however, for the systems personnel or in the case of the home computer, a parent to review and see the sites that were not allowed to pass through and add them or allow them (in case of a false positive) through.

"The self-regulatory nature of the Web community, coupled with the ease of making information available on the Web, has led some individuals to abuse their freedom of expression by putting up harmful materials on the Web. These include violence, gambling, drugs and pornographic materials." (Lee, Hui and Fong, 2003)

The article by Lee, Hui and fong identify four different methods of web filtering that are commonly used.

(1) Platform for Internet content selection (PICS): The World Wide Web Consortium (W3C) has defined a set of specifications to rate the different content on the web. This allows the web publishers to assign or associate labels or meta-data with certain web pages to limit their content from reaching users that are not targeted. The drawback of this approach is that PICS is not regulated; self-policing is required. Often, publishers misrepresent or assign labels that are inaccurate and misleading. PICS is therefore considered a supplementary tool in any web filtering system that is used in schools, universities and libraries.

(2) Uniform resource locator (URL) blocking: Certain web pages are blocked by comparing the URL (and equivalent IP address) with those in a reference URL list that is generally created by the entity attempting to block the website. This list can either be related to sites that "should be blocked" or sites that "should not be blocked." This is a very effective method of blocking sites as the list of sites to be blocked have already been identified and reviewed. The disadvantage to this system however, is that it is very laborious and time-consuming. The World Wide Web has been growing exponentially since it was first introduced to the public. It is practically impossible to keep track of all new websites that are created and launched. Reviewing the list at periodic interval to allow the list to be updated can help the process but it is still considered time and cost prohibitive.

(3) Keyword filtering: The filtering process works through the identification of key words or phrases that appear within the web content. The entire content of the web sites is set through a word or phrase check as appearing in a dictionary of "of prohibited words and phrases. If the number of matches has reached a predefined threshold, access to the Web page is blocked." This method can be very effective as some level of control can be brought to the level of Internet usage. The accuracy of this system is however, dependent on the effectiveness of identifying keywords and phrases. In addition, image content and pictorial representations are hard to classify. The most significant issues faced are over-blocking and under-blocking (these issues are discussed in detail in the following paragraphs)

(4) Intelligent content analysis: an attempt to achieve "semantic understanding of the context in which certain keywords appear" is also another method of filtering web content. "In particular, intelligent classification techniques can be used to categorize Web pages into different groups (e.g. pornographic and non-pornographic) according to the statistical occurrence of sets of features. Statistical methods such as K-nearest neighbor (KNN) classification, linear least squares fit (LLSF), linear discriminant analysis (LDA) and naive Bayes (NB) probabilistic classification have been introduced in this field of research." (Lee, Hui and Fong, 2003) The limitation of this method is that any statistical implementation is subject to noise and variation. Over time, and with continuous use, the variables that affect the noise can be significantly reduced and managed.

Very often, schools, universities, libraries and ISP providers that offer user end blocking facilities use some combination of the last three methods that have been described above to filter Internet content. The filtering process is a dynamic process that goes through successive evolutions and modifications with additions and eliminations based on the current findings and the need of the user. "A Web page is parsed and the contents in various locations such as the title of the Web page, warning message block, meta-data contents of "description" and "keywords," and image tool tips, are extracted as the features to represent the Web page." (Lee, Hui and Fong, 2003)

There are two types of errors that occur with keyword filtering software that is commonly used;

1. Under-blocking occurs when content is not blocked that should be restricted.

2. Over-blocking occurs when content is blocked that should not have been restricted.

Simply defined, these two errors occur when the filtering software 'let-through' sites that are objectionable as a result of the software not have the necessary intelligence to separate the different meaning or content of similar words -- false negatives. The second case is however, more common as the filters even legitimate uses of the words in websites therefore blocking all websites with the words -- false positives. For example, a block to filter out sites with the word "breast" most commonly found in pornographic sites also results in the blocking of valuable sites such as Breast cancer and Breast feeding -- both very legitimate sites that offer valuable information to women.

It is essential that in the process of utilizing filtering software for website, constant reviews are undertaken both manually and automatically to identify the issues of over blocking or under-blocking of the websites. It is important that when selecting filtering software schools and universities evaluate the error rate of the software being considered for implementation and the necessary software modifications that are essential in the process of ensuring that the errors do not repeat themselves once they are detected and the algorithm modified for this purpose. After implementation, the random testing of the filtering software against different urls can also ensure that the process is constantly modified and tweaked to ensure that the errors are rectified. It is also vital that the process is reliable and repeatable once the error has been rectified and the change in the filtering content made.

Values and the personal biases of the school or the individual undertaking the blocking also determines the extent of blocking that is undertaken. Often schools and universities are divided on the topic of how much blocking is desirable and the infringing of the personal rights of the students in the university. Parents however can use more stricter blocking and filtering of sites based on their own personal opinions and values. ISP are also allowing password controlled blocking that helps to control different contents and levels of filtration as desired by the adult.

Internet regulation and control is a touchy topic at best. The lack of consensus on what constitute obscene content in a website is the primary reasons why Internet censorship is difficult to define and implement. (White, 2004)

In the words of Potter Steward, the U.S. Supreme Court justice "I know it when I see it," defining what is obscene and unfit for viewing by children and young adults has been difficult to apply uniformly. Another significant and salient point of Internet censorship and filtering controls is the widely differing local laws and regulations that can make rules and regulations differ significantly between states and local communities based on the lawmakers' tolerance for policy making.

The U.S. government realizing the significance of using filters is impressing upon public libraries and schools to utilize some form of filtering software. However balancing the first amendment rights against that of the greater good of the public is a difficult task. (Swartz, 2003)

Many filtering softwares that are available are generally classified for personal use or for community use. When software is used for filtering needs such as schools and colleges the identifying the needs of all the users can be challenging. With personal filtering software, individual users can define more controlled checks over the content that is viewed. Information filtering can also help reduce the time that it takes to access the material that is being sought on the websites. (Kuflik, Shapira and Shoval, 2003)

The filtering process that is generally used is content-based filtering profiles that help the users determine the scope and the range of the filters that are needed for the purpose of the school or the university.

Content filtering is used for more than just on Internet sites in recent times. There is also an increased use of filtering content for emails and other personal correspondence through the Internet. Different filtering rules can be used based on the content clarity and the level of filtering that is needed. Often, schools and universities also offer their students free email services. This requires the university to support a wide range and capabilities of networking infrastructure. Email content and attachments are also areas where filtering is increasingly being used by schools and universities to control the content that enters their network and that can be propagated from the network. Networks are also highly prone to viruses and worms that can be easily transmitted through corrupt attachments that are sent through emails using the intranet systems that are supported by the network. The low cost and the relative ease through which the email facility can be used have also encouraged direct marketers to flood thousands of mailboxes with unsolicited email messages. Controlling this spam mail is becoming an important factor for public libraries and universities that are increasingly being subjected to more volume of mail than what the network had been initially designed for. (Sakkis et al., 2003)

Content filtering tools generally act as a block between a PC and the Internet when used. There are two ways of filtering content -- filtering software and hardware. Software filtering especially those that operate in a dynamic fashion requires a dedicated server. Hardware, on the other hand, is easier to install; although it is comparatively more expensive than the software options. (Bannan, 2001)

There are however, newer forms of technology that are more difficult to control. For example, text messaging has "the file-transfer option that can allow potentially malicious or objectionable content to circumvent normal filters." Most commercially used filtering software has a blocking effect of nearly 85% with respect to objectionable content. Schools and universities, to ensure that higher levels of security are maintained within their networks, can customize a number of commercially available options. There is also an increasing trend of using software systems with artificial intelligence for modeling and determining the extent and the scope of the filters desired. "The modeling of user preferences, content/resources and behavioral/community patterns, has led to successful systems that combine both content-based and collaborative-based methods for information filtering and retrieval." (Delgado and Ishii, 2001)

Using rating matrixes to identify the rating of websites and building capabilities within the system to constantly update these matrixes is also essential. The rating can be used as guidelines for establishing the parsing that can then take place as a next level of filtering. It is always prudent to sett up the filtering process with some baseline that can help establish the filtering process for the system. Assigning appropriate 'weights' for the system can help the process gain more confidence with making fewer 'mistakes' in the process. This method is generally used for online learning systems that is referred to as computational learning theory (COLT) or sometimes referred to as a continuous and interactive process.

Collaborating filters are based on rating or preferences of other similar category or demographics. Collaborating filters require dense data sets and are more reliable only after a significant segment of raters have evaluated and reviewed the websites. "Operationally, these methods predict a person's preferences as a linear, weighted combination of other people's preferences." (Ansari, Essegaier and Kohli, 2000)

In the case of collaborative filters, the people who do not provide preference information do not help in the filtering process for the next group of people. Content filtering however, makes recommendations based on the consumer's or end users preference for product attribute. Simply put, preferences can be identified through a series of attributes and values and often these preferences can be added to or removed based on the needs and requirements.

There are generally five information sources that can be used for the recommendations needed for collaborating or content-based filters. These are:

1. "A person's expressed preferences or choices among alternative products,

2. Preferences for product attributes,

3. Other people's preferences or choices,

4. Expert judgments, and

5. Individual characteristics that may predict preferences." (Ansari, Essegaier and Kohli, 2000)

'Memory-based algorithms calculate similarity solely based on the data available, and it considers relative contributions of the votes." (Delgado and Ishii, 2001) Computational experts are quick to identify that there are significant benefits to each systems and greater benefits if the systems can be used in sequential order to gain greater effectiveness of the filtering process.

Educational institutes attempting to implement some form of filtering processes have to be able to constant dedicate resources in the form of manpower and time to ensure that the filters being used are the most current and accurate based on the requirements. Depending on the needs of the organization, the architecture of the networks and the data basing requirements might vary significantly. The end-user of the filtration project is also an important determinant of the scope and the range of the filtration that might be undertaken. With the learning capability of the artificial neural networks (NN), it is however possible to 'train' the system to automatically identify and review objectionable websites, store these is memory for validation by a technician and then add it to the list for blocking sites.

Due to the time and the resource needs for ensuring that effective blocking of relevant sites is being undertaken, many similar agencies are also collectively working to identify site that are objectionable. (Herlocker, Konstan and Riedl, 2002)

Sites identified by one school or college is shared. And thus, a more comprehensive list can be created in a relatively shorter period of time. Human beings are able to analyze and review systems based on a wide range and scope of definitions. The same is not true for algorithms that might not be able to comprehend the other 'meaning' or inferences of terms or meanings within a text context. Visual content is even more challenging as differentiating between art and explicit objectionable images can be challenging.

Many critics of libraries using the filtering software believe that the software used is flawed. They claim that the filtering software "focus on the interest of speakers rather than the interest of their audiences, and that the interests of recipients of information are paramount in the public setting." (Bell, 2001)

Content security software can also infringe on the first amendment rights of the individual when the filtering software is sued to scan personal emails and when questionable messages or embedded images exist. Many systems are designed to automatically notify systems managers when policies are violated. (Wakefield, 2004)

In addition to blocking sites, restriction of site access through search engines is also essential. Website strategies also include the importance of being visually uncluttered and complicated. The problem of too many pop ups (or pop-unders) can also cause irritation and annoyance for the web user. The company Alexa.com markets a free tool that embeds in the browser. When a pop-up appears it allows the user the option of blocking it. While Internet advertisement and promotions are becoming increasingly common and many free sites exist by pure virtue of the advertisement, companies should draw a line at what they feel is a reasonable amount of interference with their websites.

Universities and schools are also becoming the target of spam and unsolicited mail that slows down and cripples the networking service that is provided. Denial of service (DoS) is becoming a common problem that requires schools and universities to dedicate valuable resources and time to the purpose of safeguarding the systems. Content filtering is seen as a good alternative to control the problems that are being faced. Introducing a platform for content evaluation and benchmarking the content before it even enters the system has helped control and even reduce the problems and issues of overload that universities now face. By supporting an addition platform at the perimeter of the network system, systems specialist can control the level of information flow and the quality of information flow through the system. A number of commercial off-the-shelf software packages are also available for the purpose of blocking or filtering websites, but their adequacy and effectiveness is significantly limited.

Filtering software is also being used for controlling and monitoring the use of computers on the school or university labs. It is often observed that significant resources are spent in the process of using computers for games, chatting and online auctions. "Use-filtering software is a centralized administration tool that detects the entry, presence and launch of files and applications on the network. It can prevent application installation with unique write-based detection technology. For example, it can detect and stop the installation of spy ware before it is written to the network, which strengthens security and reduces the help desk burden involved in spy ware cleanup." (O'Donnell, 2005)

It is critical that universities, schools and colleges realize that adding a content filter to their systems does not delete or alter any system files or database on the network. Rather, it acts as a layer of protection between the university/library and the Internet. One must also recognize that what is objectionable to one person may be perfectly acceptable to another person. For example, firearms websites might be considered perfectly normal for a society that is allowed to carry guns and weapons but completely inappropriate for a society that does not allow such behavior. Governing the scope and the range of content filters can be challenging as the needs of the school or the department itself within the school might require access to certain objectionable sites as a part of the study program. False-positives within the audit content is an issue that has to be investigated and rectified in the shortest possible time.

The governments in many countries are also mandating that schools offer their student's access to the Internet while at the same time ensuring that the sufficient protections has been offered to ensure that they are not accessing inappropriate websites. In the U.S., "beginning June 30, 2002 all schools and libraries are required to be following the rules and regulations of CIPA in order to continue receiving E-Rate money." (Beguette et al., 2002) The ability of schools to restrict and control the content that is accessed in schools was affirmed in Bethel School District No. 403 v. Frasier (1986). The content of the student's speech has sexual content that was considered objectionable; it was restricted. This by no means was a blow to free speech. Rather, was the extent of authority that a school could impose on the content of information that was accessed and distributed within its premises.

A Supreme Court ruling also affirmed the fact that the rights of children are not the same as the rights of adults. Responsible adults in charge of the well being of a minor had the authority to make the necessary decision to safeguard the mental and the emotional state of the child or teenager in their care. This 'responsible adult' could be a parent, teacher or individuals in the school systems that are accountable for the well being of the child in their care. The guidelines and the interpretation of 'harmful content' are vague. It does offer the individual or the committee in charge of the filter or blocking to have their say in the level of blocking or filtering that is to be undertaken. The limitations of introducing filtering content in schools are that students could obtain questionable material from their home computers and then install or load these articles or information into the school system through other methods such as text downloads.

Blocking and filtering are also becoming major issues in regions where tyrannical governments might be using these methods to restrict and limit the information that the citizens of the region might seek. As this study is only concerned with schools and educational institutions, this aspect of website blocking and filtering is not considered. Filtering of objectionable content in school is considered essential in almost all countries that have freedom of speech -- a law similar to the first amendment that is enjoyed in the U.S. It is important that filters are not used to favor one sociopolitical viewpoint over another.

"To prevent filters from being used to exclude viewpoints opposed by the software designers or other third parties, the First Amendment should be read to require that libraries using filters meet three standards. Libraries should have to (1) retain final say over selection decisions, (2) understand the criteria that the filter software uses to exclude content, and (3) have the resources to correct the viewpoint discrimination that the filters are likely to generate." (Nadel, 2000)

The judgment to block or filter content into the school is dependent on the ability of the school to arrive at the most appropriate method for determining the blocking that is required. The idea of blocking or filtering websites should not be made with the intention of restricting and controlling the information that students can gain. Rather, it should be based on what the society or the school thinks would be a bad influence on the child in their care.

When resources are limited and if the school feels that the network is being used to support activities that are not related to schoolwork it does have the right to block certain sites. This is often done so that the memory required to support the network system is available as and when needed for schoolwork. Any technology strategy is shaped by the generative forces of technology evolution and the firm's strategic action, and by the integrative, or selective, forces of industry context and the firm's organizational context." (Burgelman, Maidique and Wheelwright, 1996)

Technology strategies and the marketing of the technology are also dependent on the nature of the market and the demand for the product in the market. The same is also true for a school where the technology might be implemented for specific needs and misuse could be cost to the school over time.

It is difficult to identify a school that is not involved or needs the help of computers. Generation X is reputedly more global, technologically oriented, and culturally diverse than the generations before them. While the modern generation (Generation X, Y) has grown up on computers, older generations had to experience the learning curve for new systems. "The ready availability of a technology does not guarantee its immediate take-up and exploitation, and unless it promises some personal advantage it will not change well-established practices." (Herman, 2001)

Only when adult users view IT as a means to an end will effective use of the technology occur.

Prior to undertaking any filtering process for blocking websites, schools should step back and identify the objectives and the goals of this decision. Among the questions that need to be answered are, who makes the decision on what sites are to be blocked or accessed. (Beguette et al., 2002) In this case, involving parents and obtaining the support (at least a general consensus) on the sites that need to be blocked can help the process. When parents are also made aware of the objectionable sites and the contents that are to be blocked they can take similar measures on their home computers to ensure that access and transfer of the information from the home computer to the school computer does not occur. In turn, parents who are cognizant of new objectionable websites can also inform the school and in this manner lesser time and resource can be spent on each attempting to identify the same set of sites to block. The frequency and the collaboration that is obtained between parents and the school system will also determine the extent of interactive control that can be established.

As any systems administrator will testify, maintaining and overseeing the network system for any organization can be a challenge. The same is also true in a school. After the decision to block certain sites, schools have to be able to dedicate trained personnel for this task. In many cases, this expertise might not always exist in house in the school and an external consultant might be sought to achieve this task. In many cases, schools within a school district could also build this capability at a centralized location and offer the network services to all the schools in the area. Some level of standardization could be thus achieved among all the schools in the region. Technology conceptualization, creation, adaptation and implementation can be challenging and require dedication and focus. In addition to input from software and hardware professionals, input from the teachers and in many cases even the students can help the process of implementation. "Of course, information systems cannot be expected to serve every need" but neither should the processes required be ineffectual or cumbersome for the end user. (Petrides, McClelland and Nodine, 2004)

End users determine the extent and the scope of the adaptation of the new technology that takes place. Identifying the end user, determining his or her realistic needs and ensuring that the project results matches expectation or requirements is essential. In schools the student is the end user and identifying the needs of this user within the scope of the educational program is very important.

In reality, with any IT project of this nature, determining the true cost of the task is difficult. The main aim in any filtering and blocking undertaken by the school should be to offer students the most access to the knowledge and the information that they seek. Any knowledge transfer should be able to facilitate "knowledgeable action" from the individual transferring the knowledge and the individual receiving the knowledge. (Walsham, 2002)

Successful network systems in schools ensure that the student always has the access to the necessary source within the shortest period of time and with the least inconvenience.

Often, in addition to school personnel and parents, external agents might also be involved in the decision making process. Arriving at a consensus is always a challenge. In most cases, when a group or team of individuals is responsible for determining the best course of action for a task they employ different methods at arriving at a decision. Most teams generally utilize five levels of decision-making: command, consult, majority, consensus and unanimity. It is easy to understand that these decision types are very closely related to the time factors teams may have in arriving at a decision. When the command type of decision-making process is used, the team leader identifies the tasks at hand and identifies a course of action. Unanimity, on the other hand, ensures that all members like and accept the decision -- they have to "buy-in" to the decision. Consensus is the next option; it takes time to generate consensus for any process. In consensus, the team members discuss the pros and cons of any issue extensively. A decision is made based on the discussion; every team member might not agree to it, however. Not agreeing "with" the decision is okay; not supporting the tasks and functions that need to be completed once the decision is made however, is damaging to the team.

Different people involved in the process might also have different requirements from the filtering and blocking that they seek. It is essential to ensure that the content being blocked is 'objectionable' only with respect to pornographic, violent or related to drugs, suicide, arms and related topics. Sites should not be blocked because of social, political or ideological reasons. At all times, the First Amendment rights of the students and the school personnel should be considered. While it is true that children do not enjoy the same level of first amendment rights that adults do, ensuring that the education and the school system offer them the most complete and comprehensive education and all viewpoints on a topic is essential.

The Internet is a dynamic entity. Changes and evolution are a norm in this arena. Parents and school officials have to sport sufficient flexibility to change their decisions based on the need of the time. For example, the school might decide to add new sites to the database of sites to be blocked. This task is not a pure science and the number of objectionable sites could increase significantly over time. It is also essential that the team of teachers, parents and administrators meet to review the past blocked sites and familiarize themselves with the content that the students might be attempting to access either intentionally or unintentionally. This allows the parents and school officials to improve their own knowledge of the sites and the topics that the students are accessing.

Selection of the best vendor to provide the software and the hardware requirements for the filtering and the blocking IT systems is also a critical factor. There are a wide number of software solutions that are available in addition to the numerous hardware options. Software and hardware have their advantages and shortcomings. School officials have to be able to understand these shortcomings and limitations early on. Some vendors offer customizations and add-ons that can help incorporate future growth and development. Schools should however realize that any customization comes at a cost. Understanding the true and hidden costs of customization of hardware and software early on is essential. In reality, often, vendors selected by the school might fail to deliver on the promises during negotiations. This is not because vendors are incapable of delivering the necessary services; rather, it is more often the result of the vendors not comprehending the strategy of the school.

Successful IT projects identify the current needs as well as the potential future needs of the school. Not building sufficient capacity into the networks and the infrastructure is often cited as the main cause for many problems encountered. In addition, the range and the extent of the computer usage by the students can impact limited capacity. For example, computer games and online gaming can utilize significant network capabilities and determining if blocking the site or building greater capacity is a better option are but some of the few decisions that need to be made. "Most clients report their frustration with endless cost-service debates, and sometimes significant loss of control over their IT destiny and knowledge base." (Willcocks et al., 2004)

Vendor stability and dependability is also essential. During the pre-implementation stages, care should be taken that the hidden cost of the IT implementation are investigated. Budgets need to be planned reviewed by the relevant individuals.

Then there is also the factor of letting the student know that the site was blocked and that access was denied is essential. Students should be made aware that some form of filtering has been implemented in the school and the blocked websites are logged and archived for review at some later stage. Schools are also increasingly requiring students to use passwords to log on to computer and use the network facility. Tracking password and student ID with respect to the web content accessed can also help officials determine the trend of browsing that is undertaken by specific students. There are also other concerns such as viruses and worms that need to be blocked. These can come through attachments and Spam mail that students might receive over the school's network system.

If the school already has some form of filtering in place, identifying the shortcoming of that system can help them arrive at a more suitable new system for the school. Often, the effort to connect legacy systems that the school supports into the new system can increase the cost and the decrease the performance of the system. Adequate training and development of the school personnel to handle routine maintenance and upgrading issues is also essential. As with any IT system, training and educating the people who will be operating the system and monitoring it are essential. This is especially relevant as the current teenager, growing up with technology and having the time and the opportunity is more computer savvy that the previous generations. The ability to hack into systems and change code requirements is a real threat. It is vital that individuals overseeing the system familiarize themselves with the new methods of hacking that can be used and make the system tamperproof.

In many situations, external entities that sponsor or support computer labs in schools might also gain information through the monitoring of the sites that student's access. Data- mining and knowledge acquisition in this manner can offer these entities or school an insight into the preferences and the biases that students display. There is an increasing trend of using marketing options that can be streamlined with the knowledge that can be gained through this method. At the same point in time, students in these labs need to realize that their actions are monitored and could be documented.

Identifying the successes and the failures along the way can help the school catch mishaps and deviances from the original strategy early on and this can help them make the necessary changes as soon as possible to the strategies being implemented. If school use a centralized blocking system, appointing an individual who has the expert knowledge in this arena can help the school keep abreast of all the changes that are essential.

Language and the culture of the school are also essential factors when considering filtering and blocking. High schools in America are becoming increasingly bilingual as a result of the sizeable Hispanic population in the country. In addition, students with multiple language skills can also access sites in these languages for which filtering and blocking might not be undertaken. For example, porn sites in Spanish, Chinese or French could be accessed if the student has the ability to speak and write the language. Combating this issue of language and blocking of websites can be tricky. It is not cost-effective to block based on all languages but schools can identify languages that are spoken but their student demographics and gain insight into words, phrases and terms along with the urls of the websites that they wish to block. Flexible translation software programs can help combat this issue. "Translation between various dialects and jargons, though difficult, should still take less effort than the translation between different natural languages, since only a part of message semantics has to be processed." (Chislenko, 1995)

As translating experts will testify, software can help with the basic translations but when used for complex phrases and sentence structure these software algorithms might not be as effective as filtering or blocking the necessary files or web pages.

Very often, IT projects failures are due to the limited attention to details during the initial stages. This is also true of filtering and blocking endeavors undertaken by schools and colleges. In reality, adapting software and hardware needs to topics and conditions that are problematic only end up generating more problems and difficulties in these arenas for the school. Concerns that exist with any process or task should be first identified and solutions for these problems sought. Often, schools realize that the process of seeking solutions for problems help them gain better insights into the issues and the concerns that the process face.

All network systems that are used undergo some form of attack or intentional disaster. This is very true for school systems where students are constantly attempting to find loopholes and weak links in the system. In the case of modifying filtering criterion and blocking objectionable sites, students who are determined are constantly seeking out new ways to 'fool' the system and allow access to the websites. Understanding the risk associated with the various systems and the potential failures that could result are topics that have to be considered by any school. A centralized IT management responsible for the different existing systems and many in-house and homegrown systems support existing operations frequently does exists. Individuals who had initially conceptualized and designed the systems for filtration might have moved on to other jobs and the existing systems could be seriously jeopardized by new personnel who are not aware of the complete operations and the needs of the system.

Communication is essential during pre-implementation and implementation stage. Information that is critical to the success of the project should be transferred to the relevant parties within the shortest possible time. Any filtering system should also communicate the status of the operation to the key individual monitoring the system. Parents and "responsible adults" who are also involved in the process of constantly reviewing content filters that are used should have access to the information as well. One of the salient negative effects of integrated systems for reporting is the information-overload. Information that is communicated should be relevant and critical, while at the same time it should target the correct individuals. For example, if a breach in the filtering software occurs, the systems analyst should be the first to be notified. Similarly when decisions to update the criterion are required all the relevant parties should be notified within the shortest possible time. Conceptually, information is a critical 'resource' for the school and as such should be managed just like any resource of time, money labor. Misuse of this resource can impact other resources.

Information has to be constantly and regularly updated into the system to ensure that it is available and current. Failure to update systems makes the system ineffective and integration is never really possible. Experts in the IT arena should not overwhelm employees with limited knowledge, similarly individuals in the school system not completely aware of the complexities of the task should not belittle or minimize the true extent of the work that is required to operate a blocking and a filtering network for the school. Filtering and blocking of sites is never complete and it requires budgeting and planning just like any other department in the school. Ensuring that decision makers are aware of the needs financial and skill level can help. School's decision makers also have to use their best judgment when deciding which sites to block and the resources that can be allocated for this purpose. (Nadel, 2000)

Developing a global network with the capabilities of developing a 'brain' and creating an extensive pattern of weighted links is also essential. (Heylighen, 1999)

Students also increasingly rely on the Internet to provide them with social contacts and interactions. Emails, both solicited and unsolicited, are on the rise. While it is relatively cheap to send emails and receive them, the supporting technology requirements are is tremendous. Automated and semi-automated filtering solutions are sought to control this problem of emails that schools and universities have to deal with. (Cranor and LaMacchia, 1998)

Software filtration that is currently available is also useful for blocking more than just objectionable websites. Schools are in possession of thousands of personal records and using filtering and blocking software can also help control the intrusion of hackers from outside the network attempting to break in and steal information on students and their preferences for specific sites. (Messmer, 2005)

Many websites launch cookies on individual computers and some of these cookies could track the activities of the computer user and transmit this information. Schools are also using filters to control the online purchasing that students might be undertaking using their own credit cards or numbers that they have obtained from other sources. Web filters also are expected to become more flexible so that network managers can offer limited access to some sites based on the need of the student. For example, students studying topics on violence in the workplace or domestic violence could be allowed to gain access to sites that might be otherwise considered too graphic for the sensibilities of the student.

The amount of money being spent by corporations and schools for web and email filtering has grown exponential. In 2003, approximately $731.1 million dollars was spent on acquiring some form of web filtering software. A year later, this amount jumped to $996.6 million, approximately $200 million more than in the past year. This value is only expected to increase in the coming years. In spite of the increase of filtering and blocking of websites that is being done the topic is extremely sensitive. Schools and educational institutes have to constantly be able to justify the use of these measures for sites other than porn.

Filtering allows the imposing of values at the user end of the Internet and allows parents and policy makers greater control based on their own personal beliefs and values. Technology has offered schools a cheap and effect means for data processing, data accessibility, data warehousing and data mining. "The advancement in data processing has greatly improved the accuracy and speed at which data are entered, stored, exchanged, manipulated and analyzed." (Jain, 2003)

Schools that have the technical capabilities are also increasing using data mining to gain information on rating of websites and the content that exists on the different web pages. By running queries on search engines to identify sites that have the content to be blocked schools can constantly add the urls and new content in the form of word or phrases to the 'check lists' that they maintain.

There are important distinctions between data, information and knowledge. Data are raw facts collected by observation or monitoring; when data are filtered out to identify trends and organized it converts to information and when this information is used in the operation, planning and strategy it is converted to knowledge. Schools are also sharing this information more readily with other schools in an effort to reduce the amount of time and resources that need to be spent on the process. "Data ( Information ( Understanding ( Knowledge" is how Chowdhury identifies the use of data. (Chowdhury, 2003)

Countries and governments such as China that have used filtering tools to control the Internet use in their countries do not use any one single tool or method for filtering but rather use a wide array of filtering tools in conjunction with each other. The U.S. legal system has also been very fuzzy on their opinion of filtering and blocking of objectionable sites for minors. The "Child Online Protection Act (COPA) imposes penalties of $50,000 and six months imprisonment for knowingly posting on the Web, for commercial purposes, material that is harmful to minors." (O'Regan-Jr. And Snyder, 2004)

This has made the effort of policing the end user section of the web very difficult in countries where civil liberties and the rights of the individual are valued. The American Civil Liberties Union (A.C.L.U.) has also been unable to maintain its stand on any one side of the issue of filtering. "When a parent installs a filter that keeps a kid from seeing a bunch of sites that may or may not be pornography, that's parenting, When a government forces all adults and minors to use filters, that's censorship." (Schwartz, 2004)

It is currently estimated that approximately 40% of all the pornographic material that is posted on the web come from publishers outside the U.S. Despite many heated debates, lawmakers in the U.S. have been unable to arrive at a consensus that can help determine a law against Internet pornography. (Savage, 2004)

The unchecked legal restrictions on the Internet will only become void if the government of the United States can prove beyond any doubt that there is no other safer or more efficient way of protecting children or minors from the objectionable content of the Internet. At present software filters at home and schools are probably the most effective way of protecting minors from web content that is not appropriate. Many commercial software companies have therefore been aggressively pursuing the market in order to be able to provide a product that can be use to filter the Internet based on a set of predefined guidelines that can be imposed.

"Content filters do not scan attachments for specific viral code. Instead, they thoroughly inspect all e-mail traffic, looking for keywords, phrases, attachment names, scripting commands, and even message origin. A network manager can create rules that delineate which e-mail can pass and which should be rejected." (Mamaghani, 2002)

Commercially available software can be adequately modified to support the needs of the home computer users as well as schools if they are used in conjunction with other filtering and blocking tools.

Software development in almost all fields has been greatly supplementing the use of content filters and the effectiveness of the filters that are used commercially. Some schools and colleges that have the internal expertise to generate the necessary software systems to support an independent filtering system are able to customize the needs of the school and the extent of output that they desire from the system. Over time the information that is collected and the knowledge that is gained can be used to develop systems that are more knowledgeable and have the ability to collect, review, process and parse the websites based on broad criterion set. The increased used of fuzzy logic concepts to defining filtering needs has been observed. (Bowen et al., 1992)