July 24, 2024


Bring Out Techno

Web 3.0 & the Power of the Semantic Web

Web 3.0 describes the next evolution of the World Wide Web. The term “Web 3.0” has been floating around for quite a while now, but what does it exactly mean? The true definition of Web 3.0 has been unclear ever since it has been created. When trying to get a hold of the concept it is important to keep in mind that Web 3.0 is rather a buzzword and doesn’t clearly define anything.

The semantic web

The term Web 3.0 is very often mentioned together with “the semantic web”. Some people refer to the semantic web as an equal to Web 3.0, others consider the semantic web part of Web 3.0. What does “the semantic web” mean though?

A clever example in an attempt to make sense of the semantic web has been made by Peter Berkel. Take the words Paris Hilton; is there actually anybody who’s thinking of the Hilton Hotel in Paris? The meaning (semantics) of these two words can be interpreted in multiple ways.

In order to understand semantics on the web we also need to look at the syntax. If we use the sentence “Mike loves Kate”, the syntax refers to the composition and structure of the sentence. The semantics indicate that Mike really likes Kate. If we change “loves” into a heart symbol, the syntax will change but the semantics will stay the same.

In the example of Paris Hilton we see a reversed situation, in which there are multiple semantics but only one syntax.

In the communication between computers on the net, the syntax has an important role. When you request a website to show up in your browser, you’re actually doing a request to a web server. This server looks up your request and returns this in HTML format. Your browser reads the HTML syntax and translates this into a designed page. The meaning of what’s on this page will stay unknown to the computer.

The same thing happens when we search for information on the internet. A search engine searches through billions of pages to find the keywords that you requested. It will then show you a list of pages that include the keywords you were looking for. The search engine knows which keywords are in these pages, but the real content of the pages remains unknown. This could be compared to learning a parrot a couple words, which it will replicate without knowing the actual meaning of the words.

Using this analogy, the current web can be considered as a “web of documents”. These documents are all linked to each other in one way or another, but for web applications the content of these documents is undefined. The content however is what’s of interest to us. These documents are about people, history, companies, countries, etc. Right now we’re not talking about documents anymore, but about “entities” that are inside of these documents. The goal of the semantic web is to let computers understand the entities within these documents.

So the semantic web gives meaning to entities in web pages and relationships between those entities. In contrast to the current web, this would be a web of entities.

The benefits of the semantic web

So now we got to a better understanding of the semantic web, how can we benefit from this? Take the example of Paris Hilton. If a search engine would understand that I’m trying to find a Hilton Hotel in Paris, all results about the celebrity Paris Hilton could be excluded. This brings us to better, more accurate and faster results. This is still a plain and basic example. A next step would be an application that understands Hilton is a hotel in Paris, which can offer me more services. If you are searching for a Hilton Hotel in Paris, a Hilton Hotel could be found, while a flight to Paris could be booked together with a restaurant to eat at. This makes the results on your request a lot more valuable.

Maybe this sounds pretty familiar; when you book a flight you’ll be asked if you want to rent a car as well. This question however has been realized through a cooperation of multiple web applications. In the semantic web no cooperations or agreements have to be made since all information is universally available.

The semantic web will have a huge scale availability of information. Roughly, there are two different approaches that will lead to a semantic web: the bottom-up and the top-down approach.

The bottom-up approach implies adding information to all existing documents on the internet and making them understandable for web applications. This ‘information’ describes the entities in a webpage and all their relations. This can be done through RDF or Microformats.

Supporters of the top-down approach consider the bottom-up approach very unpractical. They don’t believe in a fully annotated internet. The top-down approach supports the development of applications that are able to have a better understanding of the data in documents. Think of tracking back the meaning of entities out of the context of a page, but mainly applications that can understand natural/logical language. Hakia (hakia.com) and (powerset.com) are Search Engines that do an attempt on understanding English in the way humans understand it.

Whatever approach will be used, the web will be more meaningful and more valuable. All entities will gather a cloud of attributes and relations. A danger would be an information overload. Ideally we only want to absorb information that raises our interest. The need of ‘personally tailored information’ will keep on getting more important.

Personal preferences

Personal preferences could be left under the control of the user. The APML (Attention Profile Markup Language) workgroup has been working towards this concept. Your profile in APML can be seen as a file with your personal preferences. Web applications are able to read this file which results in your information being filtered to your own interest.

An even bigger amount of profile information will be outside of the users’ control. At this very moment, organizations such as Google, Amazon and Yahoo are registering the clicking behavior of its visitors. In a semantic web this information will become way richer. A lot more is known about the people that visit their websites. Next to that, the web is becoming more pervasive. We are in contact with the internet all the time and will do so more and more; in our browser, our cell phone, our car, home devices, etc. Web applications will be able to register information everywhere. Profiles will be built full with information about yourself, without you noticing it.

Within this new relationship between the Web and its users, marketers will have to ‘slip into someone else’s conversation’. Brands will have to be relevant to the user’s situation, setting and desires. When the semantic web is realized and successful, it will create a huge load of information exchange between organizations and individuals, producing a new level of findability and knowledge sharing.