14.3 Content rating and self-determination

Rather than censoring what content is being distributed and made accessible on the Internet and WWW, the idea of content rating and self-determination is to enable users to judge the content of a Web site based on some objective criteria and to control access to the content accordingly . This idea actually conforms to the general argument of human beings being ultimately responsible for their own behaviors and activities.

Content rating is not something conceptually new. For some media, such as cinema movies, we are already accustomed to content rating, whereas for other media, such as television and the Web, the effectiveness and efficiency of content rating and self-determination schemes remain to be shown. ^[1] In either case, it must be ensured that content rating schemes can neither be circumvented nor manipulated. For all practical purposes, this turns out to be difficult.

The Platform for Internet Content Selection (PICS) is an initiative created by the industry to promote content rating and self-determination [4, 5]. Coordinated by the W3C, PICS aims at providing an infrastructure for associating labels with content. ^[2] It is value neutral in the sense that it does not specify the content of labels. It only specifies a label format and describes how the corresponding labels may be transmitted. As such, it is a platform on which content rating services and filtering software packages can actually be built. Computer systems can process PICS labels in the background, automatically shielding users from undesirable content or directing their attention to sites of particular interest.

The PICS specification provides the means to implement a content rating service. It consists of the following components :

A syntax for describing a content rating service, so that computer programs can present the service and its labels to the users.
A syntax for labels, so that computer programs can automatically process them. A label describes either a single document or a group of documents (provided by an Internet or Web site). A label may include a cryptographic hash value of the associated document or may even be digitally signed.
An embedding of labels (or lists of labels) into the RFC 822 transmission and HTML document formats. In the first case, RFC 822-style headers are used, whereas in the second case, the HTML META tag is used for embedding one or more labels in the header of an HTML document.
An extension of HTTP, so that clients can request labels to be transmitted with a document.
A query-syntax for an online database of labels (a label bureau as discussed below).

There are products that implement the PICS specifications, and the W3C maintains a list of PICS-compatible filtering products and services (i.e., client software, HTTP servers, proxy servers, label bureaus, and rating services).

In general, PICS can be used and provides support for both self-labeling (by an autonomous content provider or on-line publisher) and third-party labeling (by a label bureau):

A content provider or on-line publisher who wants to label his or her content must first choose which rating vocabulary to use. The W3C recommends the use of a vocabulary already used by others, to make it easy for Internet users to understand the corresponding labels. Again, a list of self-rating vocabularies is available, but W3C does not endorse any particular vocabulary. Typically, the content provider or on-line publisher chooses a self-labeling service, connects to the corresponding Web site, and describes the resource to be published by filling out an online questionnaire. After completing the questionnaire, the service gives the content provider or online publisher a text label in a special format, which is then inserted into the corresponding HTML file.
In addition to the self-labeling service, an independent rating agency need not get cooperation from every content provider or Web publisher whose material it labels. As with self-labeling described above, the independent labeler first needs to invent or adopt a vocabulary. The rater then uses a software tool to create labels that describe particular URLs. Instead of pasting those labels into documents, the independent rater distributes the labels through a separate server, which is called a label bureau. Filtering software will know to check at that label bureau to find the labels, much as consumers know to read particular magazines for reviews of appliances or automobiles.

Several PICS-compliant rating services are in operation today, allowing content providers and Web publishers to self-label their content. ^[3] The most important rating scheme and service is RSACi, developed by the Recreational Software Advisory Council (RSAC). ^[4] According to the information found on its Web site, RSAC ˜ ˜is an independent, nonprofit organization based in Washington, D.C, that empowers the public, especially parents, to make informed decisions about electronic media by means of an open , objective, content advisory system.

The RSACi system provides consumers with digitally signed information about the level ( ranging from 0 up to 4) of violence, nudity, sex, and offensive language in software games and Web sites. The corresponding RSACi levels are summarized in Table 14.1. Most importantly, RSACi is supported by Microsoft Internet Explorer. As illustrated in Figure 14.1, the corresponding Content Advisor can be enabled on the ˜ ˜Content tab of the Internet Options menu. If it is enabled, the user is prompted to enter a supervisor password (as illustrated in Figure 14.2). The aim of the supervisor password is to prevent children from changing the settings of the Content Advisor.

Table 14.1: The RSACi Levels for Violence, Nudity, Sex, and Language
Level	Violence Rating Descriptor	Nudity Rating Descriptor	Sex Rating Descriptor	Language Rating Descriptor
4	Rape or wanton, gratuitous violence	Frontal nudity (qualified as provocative display)	Explicit sexual acts or sex crimes	Crude, vulgar language or extreme hate speech
3	Aggressive violence or death to humans	Frontal nudity	Non-explicit sexual acts	Strong language or hate speech
2	Destruction of realistic objects	Partial nudity	Clothed sexual touching	Moderate expletives or profanity
1	Injury to human being	Revealing attire	Passionate kissing	Mild expletives
	None of the above or sports related	None of the above	None of the above or innocent kissing; romance	None of the above

Figure 14.1: Microsoft Internet Explorer s Content tab of the Internet Options menu. ( 2002 Microsoft Corporation.)

Figure 14.2: Microsoft Internet Explorer s Create Supervisor Password panel. ( 2002 Microsoft Corporation.)

The Content Advisor panel is illustrated in Figure 14.3. For all criteria of the RSACi system (i.e., violence, nudity, sex, language), the user can specify a maximum level that is acceptable. The Ratings tab, as illustrated in Figure 14.3, can be used for this purpose. In addition, there are other tabs that can be used to customize the Content Advisor.

Figure 14.3: Microsoft Internet Explorer s Content Advisor panel. ( 2002 Microsoft Corporation.)

There are several points to consider with care regarding the use of content rating and self-determination technologies, such as employed by PICS and RSACi:

Not every label is trustworthy. For example, the creator of a computer virus can very easily distribute a misleading label claiming that the software is safe. Checking for labels merely converts the question of whether to trust a piece of software to where to trust the label that is associated with it (and since both can be provided by the same person, they can be identical). One obvious solution is to use copyright protection labeling or cryptographic techniques to determine whether a document has been changed since its label was created and to ensure that the label is the work of its purported author.
Mandatory self-labeling need not lead to censorship, so long as individuals can decide which labels to ignore. Unfortunately, people may not always have the choice. As mentioned above, Singapore and China are experimenting with national firewalls that are going to implement some content blocking strategies. Nevertheless, it is fair to say that improved individual controls remove one rationale for central control but do not prevent its imposition .
Any content rating system, no matter how well conceived and executed, will tend to stifle noncommercial communication. Rating requires human time and energy; many sites of limited interest will therefore probably go unrated. Because of safety concerns, some people will block access to materials that are unrated or whose labels are untrusted. For such people, the Internet will function more like broadcasting, providing access only to sites with sufficient mass-market appeal to merit the cost of labeling.

As an added inducement to content rating, it is worthwhile to mention that some future applications may use labels for searching as well as filtering. Thus, rating a Web site s documents will make it easier both for some audiences to avoid the documents and for others to intentionally find them. Consequently, content rating is another example of a dual-use technology.

As of this writing, it is too early to say whether content rating and self-determination will be successful and successfully deployed on the marketplace . PICS and RSACi are supported by Microsoft Internet Explorer. There is, however, less strong support by other software vendors and content providers. In fact, there is hardly any Web site that has its content rated according to any scheme (not necessarily RSACi). Consequently, it is possible and more likely than not that the notion of content rating and self-determination will silently disappear in the future.

^[1] Note that in many countries (e.g., Australia) television shows have ratings.

^[2] http://www.w3.org/PICS

^[3] Contrary to that, there are hardly any independent label bureaus in operation today.

^[4] http://www.rsac.org