From innovative ideas and research directives, to release announcements and conference reviews, consult the STI Blog for everything related to semantic technologies.
Why does the semantic community focus almost entirely on the Semantic Web?
The Semantic Web came about for several reasons. First, following the amazing success of the WWW, Tim Berners-Lee and others promoted the Semantic Web as the next generation of the Web in which resources on the Web could be used by other resources without human intervention by augmenting them with meta-data. This would advance the Web from a Web of documents to a Web of information, which would then evolve to a Web of services, and a Web of things. For Web resources to interact without human intervention, the resources would have to be described with metadata so that other resources could understand the resources enough to interact with them. This was the first motivation for semantics for the Semantic Web. The second motivation was the explosive growth of resources on the Web leading to a need to deal with the scale. Again metadata and semantics were seen as providing solutions here.
The information systems and SOA world began using the Web as a computational platform. Prior to the Web enterprises would face integration between internal systems and with those of partners. The Web changed the computing world from intranets and corporate networks to the World Wide Web. Then organizations and enterprises were required to integrate their systems on a vastly different scale than ever before not only to do business with partners and customers in a more efficient way, but also to develop new opportunities across barriers now eliminated by the Web.
Enterprises and funding agencies looked for technologies and solutions to address this Web-scale integration challenge. The database and programming communities had failed over thirty years to substantially address integration. So the world turned to the Web and now emerging semantic Web community for solutions. This led to the growth of the Semantic Web community that included the old AI community with label changes, the new semantic technology community, and many others who changed terms but essentially continued the same work as they did in the past. Semantic Web-based conferences boomed in number and attendance. But industrial take up was slow, much slower than the take up of the World Wide Web. The growth and popularity of the Semantic Web has led to issues that dilute it such as being used as a cover for topics that are unrelated to the Web.
The new Semantic Web community was and is remarkably innovative regularly introducing new ideas with real demonstrations. The new idea with the fastest growth is the Linked Data movement that is growing dramatically and is becoming a sub-community of the Web. While the Web and the Semantic Web in particular drove the demand of semantic technologies, semantic technologies are not unique to the Web or the Semantic Web. If information resources on the Web can be augmented with metadata and reasoning to support precise search or discover, negotiations between the resources as to how to interact, selection of services or operations to apply, leading to automated or semi-automated interaction, then the same technologies should apply to all information-based systems whether the Web is involved or not. The web merely created a dramatic demand for semantic technologies to address the relevant issues.
The Semantic Technology community should review their direction and focus on developing solutions that are broadly applicable rather than restrict solutions to the Semantic Web. First and foremost, this is good computer science, and thus bucks the vagaries of trends and what’s hot. Second, the Web is only one component of the computing world. While it is exciting and useful to permit resources to interact automatically over the Web, this does not address the integration problem on which industry spends approximately 50% of its resources and that inhibits grow and agility. Third, the various strategic direction movements around the world are heading towards a common idea that is vaguely defined and variously named, but that includes the same pillars. The dominant name of this Next “Next Computing Environment” is The Future Internet . It includes:
Support for the Semantic Web is waning. Tim Berners-Lee, the co-creator of the Semantic Web, has shifted his interest to other areas and technologies to promote the growth of the Web. His dominant association is now with Web Science, the study of large-scale distributed information systems and their impact on society and the world, and vice versa. On a technical level he is supporting the Linked Data or Linking Open Data movement. On June 17, 2024 at Linked Data Planet in New York City, he called the Linked Open Data the “Semantic Web done right”. Another indication that the Semantic Web is waning is that the EU, which had been the largest funder of the Semantic Web, has terminated the program that funded the Semantic Web. The EU announced a reorganized its units and programs on July 2, 2008. The terms and names of the reorganized units did not mention semantics or the Semantic Web. In a related effort the EU is launching a very large initiative (i.e., 1 BN € from the EU and 1 BN € from industry). Again the Semantic Web is not yet mentioned in the Future Internet initiative. Due to the power and orthogonal applicability of Semantic Technologies the EU should consider making Semantic Technologies a Unit with applicability across all six pillars of the Future Internet.
Conclusion: In summary the Semantic Technologies community, and STI specifically, should redirect its efforts to support higher levels of automation across entities on massive scale in highly distributed networked environments, such as the Future Internet and not focus solely on the Semantic Web. In fact, the Future Internet is a misnomer. It should be the Next Generation Computing Environment which happens to be networked over the Internet which provides the essential element of universal access, but the Internet is as important as many other components of that environment.
Prior to ESWC 2008, a gathering of like-minded researchers took place: the first STI International Offsite. With nearly 100 attendees representing the majority of the STI member organizations, the offsite meeting consisted of a two-day session focusing on the STI achievements, structure, purpose, and future plans. Click here for the formal review.
Vision
Elena Simperl, Education Service Coordinator and recipient of the STI Service Coordinator of the Year award, lead the entire event which opened with a fairly intimidating forecast of our future digital world by Michael Brodie, chair of the STI Advisory Board. The preluding stream of shocking statistics - summarizing the increasing amount of information exchanged on today's Internet and predicting the exponential jump in the coming decade - laid the foundation for his fundamental point: "shift happens":
From a computing era to a problem solving era...
The new challenge is approaching the question of what people can do, rather than the old question of what computers can do. He proposed a new virtuous cycle between solutions to real problems and the capacity of computer science to provide these solutions; conclusively this begs a shift:
From Computer Science 1.0 to 2.0...
Currently, technology dominates (or more accurately, limits) solutions. The next generation of computer science must provision solutions which dominate technology. This advance is not a simple increment or extension. Rather a complete paradigm shift. Perhaps Einstein summarized it best: "We can't solve problems by using the same kind of thinking we used when we created them."
Download Michael Brodie's presentation.
Photo: Presentation of the Service Coordinator of the Year Award - (Michael Brodie and Elena Simperl) - Courtesy of Federico Facca
Revision
Rudi Studer, Research Area Coordinator, then presented Christian Bizer from the Freie Universität Berlin with the STI PhD of the Year award, which was followed by a presentation of his work. After his presentation, Michael Brodie again took the stage to lead a collective session on... 2020.
Entitled the Future Game, the session challenged the audience members to envision the role of semantic technologies in the coming decade. These predictions intentionally spanned far beyond semantic research in order to further define the operational tasks and necessary steps to be taken by STI as a whole. He conclusive challenged STI to become a digital ecosystem: a self-organizing network of organizations operating in (and extending) a digital environment that supports cooperation, knowledge sharing, and development of semantic technologies, methods, and applications.
The attendees were then further coerced into active participation with the assignment of coming up with a 3-minute elevator pitch answering the question: "What are the benefits of STI International and why should you become a member?" - click here to browse the creative results!
Philosophy Behind the Curtain
After the mad rush of recording elevator pitches, Sunday morning opened with a strict reminder that our basis is founded upon something much deeper than computer science: the philosophy of our modern world. The lights went off, the doors were shut, and Dieter Fensel gave a 90-minute multimedia lecture which no amount of coffee could properly prepare the naive attendee to absorb. Sparing no detail (Trinity's, blue and red pills included), the STI President dissected the Matrix Trilogy revealing the philosophical theories applicable to our work. He breezed from Kant to Hegel in order to provide a bigger picture of what STI seeks to accomplish: provisioning a world where the fine line between our physical and digital realities slowly begins to disappear.
STI International - Offsite 2008 - Courtesy of Federico Facca
Work Hard, Play Hard
The offsite concluded with the presentation of the elevator pitches before a judging committee of a handful of STI big wigs (Dieter Fensel, President; Raphael Völz, Roadmapping Service Team member; Mark Greaves, Advisory Board member; and John Domingue, one of the Standardization and Reference Architecture Service Coordinators), including John Domingue's insightful remarks on a retrospective look at the development of the field of semantic technologies in the last decade. Afterwards, Alexander Wahler, STI CEO, took the stage to thank those responsible for organizing the event and then it was off to the pitch.
STI International Football Champions 2008 - Courtesy of Federico Facca
Pre-European Championship fever got the best of us and so the offsite meeting came to a close with a much hyped football match between STI Innsbruck and STI International. Although quickly ejected due to the ol' double yellow, player/coach John Domingue boastfully carried the well earned trophy to the club after he and his International boys cleanly swept Innsbruck 5-2. Further proof that one institute will always struggle without the support of others.
Next Year
Conclusively, a success (which is saying a lot since I played for the losing side). Looking forward to next year's meeting. Please send me any comments or recommendations: graham DOT hench AT sti2 DOT org.
Browse the offsite photos on STI's flickr photostream or Photo Gallery. Also, we invite you to tag your STI Offsite photos on flickr ("STI International", "Off-Site 2008", "Tenerife, Spain"). Thanks!
A second abstraction from a white paper from John Domingue, one of the STI Standardization and Reference Architecture Service Coordinators, where the need for semantics is clearly addressed, articulating several challenges that can be met with a particular semantic solution. This post is a continuation of A Bleak Scenario - Role of Semantics in the Future of the Internet, Part 1 of 2.
Semantics enables the assumptions made by data providers to be made available in a machine understandable fashion. This in turn fosters the automation of certain classes of tasks. For example, finding online services suitable for a specific need or mapping between addresses represented in different formats (street number and postcode vs. full address).
Even given the emergence of semantics within the industrial and public arena a number of open research issues remain which require targeted funding. To date the overwhelming majority of research in the area partitions semantic reasoning from network related technologies for purely historical reasons. Typically, given a network-based request logical reasoners 'do their thing' and pass back results en masse. What is required is a programme of research that unifies semantic methods and techniques with network technologies at a fine-grained level. Indeed, several challenges can only be solved with a truly integrated semantic-network approach:
- Intelligent large scale content access - specifying what is desired or required in a way which captures the relevant context and is separated from the location, ownership, and technical details of the content.
- Identity - the ability to uniquely specify and identify the actors and artefacts at Internet scale is not solved. Note that uniform naming schemes are also a partial solution. The ability to interpret, for example, "The British Prime Minister" and "Gordon Brown" as the same individual is paramount.
- Scalable reasoning - the ability to reason over billions or trillions of data items, and to offer partial results in a coherent understandable manner is key.
- Mapping meanings at scale - automatically linking and mapping between large volumes of data items represented in different formats will be essential for many Internet related activities. For example, mapping between photo data (location, time and owner) for the contents of Facebook and Flickr is not feasible in an automated fashion at present.
- Trust at scale - within scenarios where large volumes of data are integrated on-the-fly from multiple sources and online brokers act as intermediaries enabling trust will require extensive research.
- Targeted research would fund a programme which integrates automated inference in the following ways:
- Inference with search - combining in an intelligent fashion the activities of locating data with creating new data through inference;
- Inference and human computation - a number of interesting recent scenarios have demonstrated that people masquerading as intelligent services ("Artificial Artificial Intelligence") can be used to resolve tasks which are currently computationally intractable. A seamless combination of logical reasoning and human computation is required in the networked age.
- Inference and social network processes - the benefits in leveraging the power evident when human social processes become visible online has been shown by the value of companies such as Facebook and Bebo. Tightly linking online social arenas to logic based reasoning will open up completely new markets and opportunities.
We invite your comments with regard to the general projective overview summarized above. Moreover, if you would like to actively take part in realizing this vision of the Future Internet, we encourage you to join the STI Community.
One of the spotlighted themes of the upcoming ICT 2008 this November is the Future Internet, a clear indication of the demand and support for innovative ideas which contribute to the development of the next generation of the Internet. Below is an abstraction from a white paper by John Domingue (The Open University), one of the STI Standardization and Reference Architecture Service Coordinators, that clearly addresses the current status, upcoming challenges, and the pertinent role of semantics in meeting the challenges confronted by the evolving Internet. Refer to The Panacea - Future of the Internet, Part 2 of 2 for further examples of several challenges that can only be solved with a truly integrated semantic-network approach.
The context for ICT research today is best captured as the 'networked age'. Network technologies are the fabric through which we now communicate, work and socialise. However, our new technological infrastructure brings challenges associated with the sheer volume of data and users. For example, in 2006 161 exabytes (108 TB) of information was created or replicated worldwide. IDC estimates a six-times growth in this metric by 2010 to 988 exabytes (a zetabyte) / year. More generically, estimates are that while currently new technical information doubles every 2 years by 2010 this will be every 72 hours. From a user perspective a similar story of large numbers and significant growth exists. In April 2007 MySpace had over 200 million registered users - making the MySpace population the 5th largest country in the world. We also know that the number of users on the Internet will grow from today's 1.1 billion to 4 billion with the addition of mobile phone owners.
A key additional aspect tied to the above statistics is that the acceleration is driving by users and not by business. Consequently, unless we enable a paradigm shift, within a few years the majority of the data available and communicated across our networks will have virtually no structure. A bleak scenario envisages a world where by-and-large the use and management of the available data requires significant labour by ICT experts. In effect users - be they generic home-users or senior personnel within a corporation - will be required to communicate through (maybe distant) technical experts in order to access, interpret or manage their data at any level of accuracy.Within the research community, semantics (the ability to encode a level of machine-understandable meaning) has been seen as the answer to the above. Semantics has been seen as the only viable option for over a decade and now a significant community exists as evidenced by the Semantic Web conference series, which now exist internationally, in Europe and in Asia and regularly attract large numbers of participants. We are now observing a major shift as we witness the emergence of semantics as a key topic within industry. There are various items of evidence for this. For example, in June last year Microsoft organised a SemGrail workshop in Redmond which focused on the provision of semantics at Internet scale. Also, in March this year Yahoo announced that they will embed semantics into their search engine. Given that some estimates put the value of 1% of the Web search market at one billion dollars; Yahoo's announcement is most noteworthy. Finally, Oracle recently released their Spatial 11g platform to support "enterprise-class semantically-enabled business applications" in a number of large business sectors.
Although extremely successful widespread recognition exists that the current Internet, designed in the 1970's, contains a number of flaws which will hinder continued growth of use. To address this a number of new initiatives have begun across the globe to design a new Internet fit-for-purpose for the networked age. Within Europe the Future Internet initiative began in March this year with the gathering of over 70 projects, with a combined value of over 500 million Euros, at The Future Internet Conference in Bled, Slovenia this past spring. Discussions at Bled have led to an initial vision for the Future Internet incorporating:
- Internet of Things - where every electronic devices are active participants in the network;
- Internet of Services - where applications live in the network and data is an active entity;
- Internet of user generated content - as mentioned above, where user-generated data is king;
- Internet of publicity, privacy and anonymity - where people and software understand how far trust can and should extend to others;
- Internet of mobility and ubiquity - where pervasive 24/7 connectivity is the norm.
Within Bled agreement was reached over a number of cross-domain issues of which scalability and interoperability (the ability to link two or more computing systems) were prominent. The existing evidence is that semantics is the only viable approach to tackle these two issues systematically. Indeed, a number of the projects represented at the Bled meeting already have semantics established within the core of their approach (such as SOA4All and Service Web 3.0).
We invite your comments with regard to the general projective overview here summarize. Moreover, if you would like to actively take part in realizing this vision of the Future Internet, we encourage you to join the STI Community.
Welcome. As with all "first" entries, we at STI International would like to take the opportunity to define the context of what we plan to post here. The STI Blog aims to provide:
We encourage the community members to comment on interesting blog posts. Join the STI Community.
We look forward to the contributions and support from the STI Members. Become a STI Member.
We also welcome any comments or complaints with regard to topics covered or opinions expressed (please contact Graham Hench, graham DOT hench AT sti2 DOT org). Looking forward to your input!