Over the course of a relatively small amount of time, the Internet has changed dramatically, going from merely collections of static corporate Web pages to a network of real-time interactions. An imminent revolution is primed to spur a shift in usage patterns and economic models. The driving factor will be a new class of easy-to-create, user-friendly electronic appliances dedicated to specific tasks, as well as devices that can analyze Web data and quickly retrieve requested information. In other words, devices are optimized according to the context in which they are used, leading to innovative products and services collectively called the Contextual Web.
The emergence of this new class of products and services is a result of the necessary existing technology and the relatively low barriers to entry in providing Internet-related software and services. Technology startups frequently engage in fierce competition to attract users by providing evolving Web-based services. Yet, many fail to generate revenue streams from a user base that is addicted to free access to data on the Web. Adding a hardware component—an information appliance—enables a complete product experience where consumers feel value in their purchase.
CONSUMERS BECOMING CONTRIBUTORS
During the initial Internet boom, information was produced by the organization, company, or individual running a given site and typically retrieved by users reading Web pages or downloading files. The Web operated as this one-way street with users consuming information made available by another party.
Several years ago, the arrival of interactive and participatory features accessible to a wide range of users—dubbed Web 2.0—signified a major shift in Web usage. Here, users can engage in two-way communication on Web pages with the ability to create and share content. Popular examples of user-driven Web sites include YouTube, Facebook, MySpace, Wikipedia, Flickr, and Digg. In the past few years, increased interactive features on the Web have dramatically changed the way people connect with family, friends, and colleagues.
While data is increasingly delivered from diverse sources, users continue to work within “walled gardens,” the Web term for independent islands of applications and information. For example, data supplied to the social network Facebook is typically isolated from data given to other social-networking sites or professional-networking sites. Users manage each site separately by entering private log-on data, having separate sets of “friends,” and accessing data one site at a time.
Attempts to unify the walled-off information thru application programming interfaces (APIs) are limited by the business models employed by service providers that wish to control the portal aspect of the user interface to generate advertising revenue. More inventive schemes place advertising a bit more transparently into the content itself, but the fundamental fight over consumers’ attention inherently creates barriers between those consumers and the information most valuable to them.
Technology enables valuable data to be freely shared between users who are also content creators. An example is OpenStreetMap.org, which allows consumers to contribute updates to maps using GPS logs and interactive sessions with their Web browser. Over time, the data produced provides a valuable resource without putting the full cost burden on any individual or institution.
An information appliance aids consumer contribution and utilization of data on the Web. This device focuses on the user experience and accelerates the creation of valuable data into information services available.
THE SEMANTIC WEB
The Semantic Web is a fundamental attempt to remove the walls from walled gardens and enable all the information to be used simultaneously. Today, an individual can visit a Web page, quickly interpret its content, and determine if it has any information of interest. While searching for common information, many users are too impatient to fully research each Web site returned in search-engine results. Instead, consumers only focus on the earliest results, limiting their chances of finding a site that meets their needs. This can be avoided, since fundamental technology is available to allow the computer to search the entire Web for user-relevant information based on pre-determined specifications.
Ideally, the next major step is allowing the Web to operate as if it were one massive application, capable of accessing any text, images, video, or other information regardless of its location. This is the ultimate dream of the Semantic Web, also referred to as Web 3.0, in which all content on the Web is accessible as if it were stored in a gigantic database. In this vision, specialized search engines, electronic devices, or appliances fully understand all content across all Web sites and are able to analyze and present all the information strewn across the Web in a tailored user-friendly fashion.
Having all Internet content put into the framework of the Semantic Web will allow applications to understand how to retrieve information using database entries, even with access control, allowing users to remove mundane tasks such as copying and pasting information. When customers synchronize and publish their photos on social Web sites like MySpace, Facebook, or Flickr, all of the appropriate recipients receive notifications and data as the customers desire without having to use a particular service. These photos integrate with their photo management applications without manually triggered downloads.
It is important to note that the Semantic Web itself is not an application or a physical device. It is content expressed using an evolving set of specifications. It enables people and devices to find and use information on the Web, but there must be a common vocabulary that exists between information warehouses and data-retrieval hardware or software. Only with consistent vocabulary will it be possible to search through an infinite set of resources, allowing an information consumer—man or machine—to focus on the important information regardless of source or type of material.
A COMMON VOCABULARY
So why is this vision of the Semantic Web yet to be realized? The difficult task is getting Web publishers and developers of information-consuming applications or devices to agree on the types of data to share and how to validate the recipient of such data.
The problem of common vocabulary is slowly being recognized. In current Web documents based on HTML, there are limited ways of expressing information to enable a machine to read it. To be more efficient, a scheme—Resource Description Framework attribute (RDFa)—has been designed to publish data only one time instead of two. With RDFa, it becomes possible to use custom vocabularies to include machine-readable data in Web documents.
Uses of RDFa are already implemented. For example, Yahoo! recently announced SearchMonkey, a set of tools with which developers can create custom ways of extracting and presenting data from Web pages. Thanks to RDFa, Yahoo! SearchMonkey understands several different vocabularies, including calendar events, contacts, addresses, locations, reviews, friends, syndicated updates, résumes, and more.
In addition to being able to enhance the visual appearance of search results, SearchMonkey enables Yahoo! Search to discover very specific content, such as résumes for people with RDFa experience located in your vicinity, without having to visit every job and social networking site. Using SearchMonkey, developers and site owners can use structured data to make Yahoo! Search results more useful and visually appealing and, thus, drive more relevant traffic to their own sites.
Common vocabulary must include not just data, but also a way to validate recipients. OAuth is a protocol that allows applications to be granted access permissions by a user without being tied to a specific service provider. This third-party validation scheme prevents the need for a separate log-in on each service used. Unlike some other schemes, OAuth gives users control in determining which machines are enabled to provide information to verify identity and authorization.
Despite the glowing promise of the Semantic Web vision, specifications such as RDFa, useful vocabularies, and third-party validation schemes such as OAuth, remaining gaps prevent the vision from reality. The Semantic Web is a solution with concerns, including the large amount of effort and time needed to recode Web pages to be compatible with RDFa. Without the common vocabulary, it leads to a bigger fundamental concern that the Semantic Web vision fails to satisfy sufficient motivation for Web publishers to make their valuable content widely sharable. There must be an underlying motivating factor, which generally means an economic incentive.
THE CONTEXTUAL WEB
The economic motivation for developing the Semantic Web results from another form of information collection and dissemination called the Contextual Web. It is an expansion of the Semantic Web, consisting of hardware and software that address usability needs under the context of specific-use cases. The standards of the Semantic Web still apply for interoperability, but meaningful extensions are provided. The goal is to create Web services and devices—information appliances—that access and generate information in a user-friendly manner.
The value-extraction conundrum imposed by the Semantic Web, where RDFa information adds value to the data hosted by a Web site while simultaneously simplifying the extraction of data without supporting advertising, can be overcome in several ways via the Contextual Web. As suggested earlier, advertising could be better nested within the content or even introduced as product placements and placed directly within the source material.
The addition of specialized hardware and use of open standards are the keys to making this vision viable. While current Internet usage is centered around the PC and, to some degree, other general-purpose devices such as PDAs or mobile phones, this new vision shifts the center of activity to an entirely new type of consumer electronic devices. To share information, there are two key requirements: network connectivity and storage. These need not be tied to a PC. Rather, they can be implemented in a convenient appliance dedicated to a specific purpose. Users will continue to pay for connectivity and storage, but with the Contextual Web, they will do so in new ways.
The Contextual Web is a combination of three things—a device, information, and a business model—all based on a Web interface. A dedicated appliance would be designed to perform a very specific task with the consumer merely needing to push a few buttons or menu selections without having to log on to the Web in the traditional manner. The machine will automatically perform all the required tasks, except collect user input and provide results, and significantly enhance the overall user experience.
With the Contextual Web, portal vendors no longer feel a need to lock out additional sources of data because they have sufficient incentive from hardware revenue. Their portal remains accessible through the Web and becomes an advertisement for their hardware that provides a richer and more efficient interface.
Existing services that do not comply with Contextual Web extensions would continue to operate, thereby providing no economic disincentive to purchasers of the Contextual Web hardware. Services that do not extend their interfaces, however, will be disadvantaged relative to the contextually optimized behavior of those who do.
A NEW WAY TO COLLECT AND PRESENT INFORMATION
This idea will lead to a new class of consumer devices devoted to specific applications and, ultimately, a simple, easy user experience. For instance, a recipe tablet might present cooks with a menu that provides a selection of appetizers, main courses, or desserts. It would then list a few ingredients and, finally, search the multitude of recipe sites on the Web and present only those that match the given criteria. The results wouldn’t return links to appropriate Web sites but instead provide the recipes themselves. This browser appliance understands that it’s a recipe tablet and only needs recipe information.
Going further, such a device could indicate where in your neighborhood you can purchase unusual ingredients or locate where discounts are being offered for the needed ingredients. To be included in these listings, stores would have to ensure the content on their Web pages is coded to be machine-readable, giving them a true economic incentive to participate in the Semantic Web.
Next, consider a wireless device, such as a mountain-hiking GPS device. It recognizes the hiker’s location and retrieves hiking maps of the immediate vicinity, suggested routes, and comments from previous hikers. It also displays nearby huts, mountaintop restaurants, the closest public transportation, or any other relevant information the appliance developer believes would give this device market appeal. All this information is readily available on the Web but would entail multiple searches. With a few button pushes, though, a mountain hiker could gain direct access to it.
Also, consider an appliance marketed toward a specific Web site. Soon, consumers could purchase a digital photo frame marked “Flickr-enabled.” After making a few touchscreen selections, the frame could search Flickr and other Web sites containing photos, download those that correspond to the menu selections to the frame’s local memory, and begin to display them. An advanced digital photo frame might further use GPS-captured data to represent the location at the time of day the photos were taken and superimpose the pictures, with the path you took, on a map.
In distributing information in a machine-readable format, standards for permissions are an important issue. Behavior is changing, and the ideas of privacy are significantly different from previous years. Society is grappling with the issue of privacy on the Internet. While users expect data to remain private, the information usually is public. For instance, if you take a photo of your family, what rights do the people in the background have? Could face-recognition software be developed to identify an unknowing individual’s location, time of picture, and who they were with while having this information published on the Web?
Customers ultimately want to control their data rather than have a Web service control it. Right now, users are sometimes required to provide personal information to a Web service. Given the choice, users would opt to not provide private information. A third-party verification scheme will be required since consumers want to be able to confirm who they are connecting with through social networking sites.
To develop the Contextual Web, developers must address a number of concerns. What kind of computing power and storage will these new devices need? What operating system will they run? How will they get network connectivity? What language will a device speak to interact with multiple services? What economic framework will make it acceptable to consumers?
However, development costs for these new systems can be cost-efficient by using standard Web technologies based on existing platforms and hardware modules. A hardware/software platform that can browse the Web with specific goals in mind will ideally become the development platform. The necessary connectivity, storage, and tools to start developing Contextual Web applications and devices are readily available. However, the largest issue remains. There must be an economic model and user experience that makes this power attractive to a large segment of consumers, independent of their level of Web sophistication.