History Of The Semantic Web Technologies
Introduction To The Semantic Web Technologies
The Semantic Web community has come a long way since its beginnings in the late 1990s and early part of the twenty-first century. There are obviously many ways of categorizing the progress of the topic; however, one easy fit is through three broad phases. Pre Semantic Web, various groups across the globe who had been working in areas related to semantics began to think how semantics and ontologies could aid in certain tasks related to the emerging Web. There then followed significant research funding, which was measured in hundreds of millions of Dollars and Euros, fueling large collaboration projects first in the USA and later in Europe.
Over the last few years, we have seen a second shift. The word "if" disappeared from the vocabulary used in relation to the Semantic Web a long time ago. Now the word "when" has gone too. Primarily around the term Web of Data, we can see commercial take-up by some major players in the area of the Web, Social Networks, and Media. Right now discussions focus more on which particular branch of Semantic Web technology will grow the fastest. We have also seen in Europe a movement away from specific large research calls in the Semantic Web area to semantics being a component within other domains, including, for example, services, media, security, and even networks. This take-up by both research and practitioner communities has been largely based on the fact that the Semantic Web is built upon global standards.
Brief History Of The Semantic Web Technologies
It is hard to know who first had the idea of creating a language on the World Wide Web that could be used to express the domain knowledge needed to improve Web applications. By the mid-1990s, before most people even knew the Web existed, several research groups were playing with the idea that if Web markup (which was all primarily HTML) contained some machine-readable "hints" to the computer, then one could do a better job of Web tasks like search, query, and faceted browsing. It is important to note that at that time, the potential power of the Web was still being debated, and there were many who were sure it would fail.
However, by 1997 or so, it was clear that the Web was going to be around for some time, and there was a burst of energy going on. Various researchers were publishing algorithms, suggesting that different approaches could be used for searching the Web rather than the traditional AI approaches, and it was around this time that Sergey Brin and Larry Page published their famous "PageRank" paper, which led to the creation of Google and the growth of the modern search engine. This historical event is mentioned here as it is sometimes said that the Semantic Web was created to improve search. This is partly true, but it is important to note that search as known back then, pre-Google, was not the same as the current keyword search that powers so much of the modern Web today.
At this time, the first "real" refereed publications were also seen coming about machine-readable knowledge on the Web. One of these approaches was the SHOE (Simple HTML Ontology Extensions) project, which took place at the University of Maryland. The slogan for the SHOE project, which continues to be a popular quote in the Semantic Web community, was "A little semantics goes a long way," and supporting this slogan the SHOE Base Ontology contained a very minimal set of concepts and relationships. Around this effort, a number of tools were created within the project including a semantic annotator for HTML pages and a semantic search tool.
Another early project was Ontobroker, which, like SHOE, looked at adding and using semantic annotations to HTML pages. These two early projects looked at what is now called Web ontology languages, and were driven less by the AI-inspired push for expressive languages, and more by the needs of the emerging Web – what would now be called semantic annotation or tagging.
Other early projects within Europe included On-To-knowledge (which began in January 2000) from which the SESAME repository was developed and also OIL was set up as a cross-project initiative, merging an effort called XOL (XML Ontology Language) and the work emerging from Ontobroker. Approximately 18 months later the OntoWeb network of excellence started, which was the birthplace for the Knowledge Web project.
In parallel with this Web representation work, W3C had begun to explore whether some sort of Web markup language could be defined to help bring data to the Web. The Metadata C Format working group was drafting a language that was later to be named the Resource Description Format (RDF). There was at this time a split between XML and RDF, which we do not have space here to recount but suffice to say that this added confusion to the overall story.
It is also worth noting here the dialogue that began in the late 1990s within the Knowledge Acquisition Workshop Series in Banff on the relationship between knowledge acquisition, modeling, and the Web. One of the projects that came out of this discussion was IBROW3, which examined how knowledge components could be reused through the Web. Elements of this project later influenced Semantic Web Services research.