Nikolaos Konstantinou of Athens Information Technology (AIT) and colleagues at the National Technical University of Athens (NTUA) in Greece state that after almost a decade of research, the fundamental concepts that would underpin a Semantic Web have matured, yet the average web user cannot yet take advantage of their full potential. They suggest that there are three main issues to be overcome before Web 3.0 emerges and they present a roadmap in their paper, "Technically Aapproaching the Semantic Web Bottleneck," to explain how these must be addressed.
The first incarnation of the web was composed of static websites that linked to each other, and search engines to help you find sites of interest. Web 2.0 brought a social element to the web — with users sharing, commenting, and interacting through sites such as YouTube, Facebook, and Flickr. The future web — the Semantic Web or Web 3.0 — will embed meaning within digital information so that any given page can be understood by computers as well as people.
The inventor of the World Wide Web, Tim Berners-Lee, first mentioned the concept of a Semantic Web, a web with inbuilt meaning, long before the advent of social sites, but it is yet to become reality. This is despite the ongoing efforts of web engineers, academics, search engine companies, and the web industry itself. There is, researchers writing in the International Journal of Web Engineering and Technology, a semantic web bottleneck.
In Berners-Lee's original vision for the Semantic Web, machine-readable information embedded in a digital object (metadata) would allow software to potentially understand the meaning and context of the digital object. Although some software currently has a limited understanding of simple metadata, this capability mostly lies in prototypes and lab environments.
However, the potential of the Semantic Web is to have software agents that can perform tasks automatically based on variables such as a user's behavior or preference settings, and to locate pertinent information far more efficiently than could an individual searching the web manually. The software might also be able to infer additional knowledge based on previously existing information processes into a usefully organized format. Such a process would be useful to scholars, doctors, engineers, scientists, musicians, designers, artists...indeed, anyone who works with data.
Konstantinou and colleagues point out that three issues are preventing this from happening: a lack of simplicity, a lack of integration with existing technologies and practices, and a lack of overall adoption by the web industry.
They suggest that ways to automatically add metadata to digital objects are now needed to make it possible to publish semantically rich content without manual intervention, regardless of whether the "publisher" is a large corporation or an individual content creator. They also say that semantic technologies should not be offered as a substitute for current practices, but rather as a complement to them, and that web engineers need not abandon experience, but should build on it. Finally, the driving forces of the web industry should adopt semantic web technologies because their adoption entails benefits both for the companies themselves as well as the end users. "This seems to be the most promising solution for the chicken-and-egg problem of the Semantic Web," the team says. "Much still needs to be done in order to effectively publish and exploit large-scale semantic information. Following the approach suggested in this paper, we are confident that the Semantic Web bottleneck will be shortly circumvented and the Semantic Web vision will be at last realized," the team concludes.
Konstantinou et al., "Technically Approaching the Semantic Web Bottleneck." Int. J. Web Engineering and Technology, 2010; 6: 83-111