The Semantic Web – Making sense of search

We’ve had the commercial web; we’ve had the interactive web. Now, the Semantic Web is apparently moving the Internet to a more meaningful level. David Bicknell reports


We are all used to the idea of ‘search’. So much so, that ‘Googling’ has even become a colloquial verb that will probably find its way into the next edition of the Oxford English Dictionary.


However, the web’s user-unfriendliness means that every search pulls up too many references, and a list of results that then requires the user to interpret raw data the computer has thrown up.


Effectively, web pages are simply not smart enough for software to understand and process the content of the page. What we need is some way of adding meaning to the data. Hence the term semantic; semantic refers to the meaning or the interpretation of a word, sentence or other language form.


It is this ‘Semantic Web’, or what some have called Web 3.0, that is exercising the minds of researchers, universities, and forward-thinking travel businesses worldwide. While Web 1.0 referred to the first generation of the commercial Internet, with content that was hardly interactive, Web 2.0 went far beyond that, with the adoption of tagging, social networks, and user-created taxonomies of content, dubbed ‘folksonomies’, such as Flickr, Del.icio.us and Wikipedia.


What Semantic Web research seeks to produce are machine-readable languages such as RDF (Resource Description Framework) – which are a consistent, standardised way of describing and querying web resources, from text pages and graphics to audio files and video clips – that allow web content to be indexed and retrieved more intelligently.


The pioneer of this new approach is World Wide Web inventor Tim Berners-Lee, now director of a web standards group, the World Wide Web Consortium (W3C). Berners-Lee, together with Wendy Hall and Nigel Shadbolt at Southampton University in the UK, and MiT in the US, are influencing Semantic Web research around the world


What the Semantic Web offers is the ability to mine web-based information that is currently trapped in isolated silos of incompatible databases, and add meaning to that intelligence, creating not a catalogue, but a guide. Ultimately, in travel terms, this would mean developing a system where a user could search for a holiday simply by typing in search terms along the lines of: “Find me somewhere warm, within five hours of flying, because I have a young child with me. And a budget of £1,500.”


Such a search could take so long, you might feel like going to a high-street  agent in the first place. Indeed, what the Semantic Web will eventually do is replicate that holiday package through the web just as well as if it had been planned by your travel agent. And it will encompass all that ‘Tripadvisor-like’ feedback from fellow travellers that your high-street agent can’t give you.


However, if you’re hoping to utilise the Semantic Web to book your next holiday, you’ll have to think again, because it is still only an evolving extension of the web we know today. Ultimately, however, web content will be expressed in a form that can be understood, interpreted and used by software agents, permitting them to find, share and integrate information more easily.


At the heart of the Semantic Web, alongside RDF, is a concept called the Web Ontology Language, which represents both the meanings of terms in vocabularies, and the relationships between those terms in a way that is suitable for processing by software. An ontology has its roots in philosophy and is the study of ‘being’ and ‘existence’.


For now, there are already companies supplying Semantic Web-based tools for today’s travel websites. One, Vienna-based Lixto, recently won an award in the annual Semantic Web Challenge competition with its Personal Publication Reader application, and its tools such as Visual Developer and Lixto Suite are already being used by European tourism portal Visiteurope.com and by the Austrian National Tourist Office.


Nilhan Jayasinghe, head of search at Spannerworks, foresees some intriguing times ahead once the Semantic Web becomes a reality.


“What will happen in the traditional advertising market? A machine doing search is not going to be biased towards or influenced by what’s been seen on TV. So there is a big implication there in terms of what will happen with marketing.”


TUI new media director Graham Donoghue describes the Semantic Web as “an advanced version of RSS on steroids”.


He believes the challenge for online travel companies will be knowing when to consider its principles in their future web design.


“You are eventually going to have to write your website using a new semantic definition, and we will start to see elements of semantic search integrated with existing travel products. But it will take time.


“For example, RSS has been around for four years, and it has only now become something of a standard for websites to offer. The Semantic Web will take off, but only when a prototype of a standard has been developed to give it momentum.”



What do they think about the Semantic Web?


Kevin Cornils, CEO at Buy.at, the UK’s largest independent affiliate network, whose travel clients include Cruise Thomas Cook and LateRooms, says: “Adding this extra information to web pages will enable precise contextual advertising – which is of interest to us as a leading affiliate network – and the monetisation that goes with it. With information more precisely indexed, Semantic Web results should appear higher in the search engines than traditional pages.


“The biggest change in the past two years is the amount of data millions of people are entering in a structured way on Web 2.0 sites such as Facebook, Myspace and Bebo. People are now adding structure to information through tags – and others are then using these tags to find that information.”


Mike Teasdale, planning director at Harvest Digital, whose clients include Thomas Cook and Cheapflights, says:
“My understanding of the Semantic Web is that data is separated from content and is easily machine-readable for search engines to strip pricing data out of holiday sites and offer detailed price comparisons of, say, the cost of a room for two in a specific hotel in Dubrovnik.


“In theory, that’s great. But my view is that agents and their websites should be offering a more subtle advice function. I don’t necessarily know that I want to go to a specific hotel in Dubrovnik. What I do want is a holiday for my family where we’ll have a lovely time and I won’t be bankrupt at the end of the fortnight!


“That kind of intuitive advice is what agents are good at, but it’s an area where computers struggle. By stripping everything down to price and availability, there’s a danger that anyone attempting to offer good quality advice will actually end up uncompetitive.”



Still confused?


The Semantic Web Revisited
A World Wide Web Consortium paper
Lixto demos 1 and 2
Social Media meets Semantic Web
The Satine Project



The Semantic Web, in brief – according to Ken Leeder, chief executive, RealTravel


1. What is the Semantic Web, in layman’s terms?
It is a web where computers can share, and make use of information, in a meaningful way.


2. Why should the online travel industry be excited about it?
Complex trip planning can be achieved by combining information from multiple sources. The Semantic Web has the potential to increase the types of functionality that a travel site has, while reducing costs.


3. When will we see it in action? And how?
We are starting to see it now. RealTravel already incorporates information from multiple sites in a unified and relevant manner, as well as making our content available through standard web service APIs to other sites.


4. What does this mean for users’ searching habits and expectations?
This means the web becomes an even more powerful travel planning tool for people and as such user expectations will rise. That enables organisations to more easily bring together the information that people need to plan the perfect trip.”

This website uses cookies to ensure you get the best experience. Learn more