So, the important point is; it is a big push to the current web to shape it more meaningful way. Why should a market dominant do it? Google search engine is powered with NLP (Natural Language Processing) and it works very well but it may be useless when compared future data on the web. NLP is a powerful technique which is aimed to translate spoken languages to machine ones to process, analyze and use it. There are many languages in the world which Google has to deal them separately and it is doing it very well. So, what is the challenge? It is social networks.

It is easy to manage your web site or web page content by yourself and ensure that it is well formed and correct for search engines. However, it is impossible to manage user distributed data over social networks. And, the easiest way to do it; add semantics to the content which will help search engines (or whatever you have) to process it.

I guess this push will improve the feature of the web and also semantic web. It seems we are going to see more tools, libraries, frameworks built in support for RDFa and microformats.

Related articles by Zemanta

• RDFa, Google (ivan-herman.name)
• Google Takes Steps Towards A More Structured Web (techcrunch.com)

I made my mind with help of my advisers and decided to study on Peer-to-Peer (P2P) Semantic Networks and made a three phase plan. First phase is supposed to finish in three weeks time. That’s why I spent most of my weekend implementing a p2p network using twisted networking framework.

It is an implementation of simple client-client p2p network which is designed to work in local networks and doesn’t support file transfer protocol at the moment. So, it is the simplest p2p network that each peer can discover and talk each other for specific tasks. It is planned to accomplish basic p2p operations to test our main semantic application.

Now, next step is to add some semantics to this p2p network, so, I’ll be working on NLTK (NLP, Natural Language Toolkit) for two weeks.

PS: Twisted GUI integration easier then cli (command line interface). Strange, isn’t it?

Related articles by Zemanta

• How the Semantic Web is Changing Search Engines (browsers.suite101.com)

Today, I felt much better when I saw the following videos. It was a right choice to make investments on semantic web. I hope, it will much better… Thanks Cas for sharing these nice videos:

Future Internet Video from Castemelijn on Vimeo

Impressed Kevin Kelly’s presentation and the quote about web: “Have to get better in believing the impossible”

Tomorrow, I have a meeting with my advisers about Semantic Web for P2P networks and if it goes successfully, I’ll start build my project in the following weeks. First, I want to up and running simplest p2p network in my local machine using twisted on python or jxta on java. Then give this thing a mining: semantic …

May be I should try it in country-wide.

Related articles by Zemanta

• How to get your feed into Zemanta? (zemanta.com)

• Ontology learning from texts
• Ontology learning from dictionary
• Ontology learning from knowledge base
• Ontology learning from semi-structured schemata
• Ontology learning from relational schemata

Almost all of the techniques above require a good knowledge and infrastructure of NLP (Natural Language Processing) to obtain effective results. I though ontology learning from dictionary would be easy compared to other methods for Turkish language. However, it requires WordNet like dictionaries that what I don’t have. Sabancı University NLP group works on BalkanNet but I don’t think it is ready to use.

Also, I’m searching about ontology-based system designs and implementations.

Reference
http://www.sti-innsbruck.at/fileadmin/documents/deliverables/Ontoweb/D1.5.pdf

My adviser already has a proposal on multimedia related systems and suggests me to consider it. I may combine those studies with ontology based infrastructure and come up with new subjects. But still an idea …

I’m listed in Semantic Web Twitter Pack as Semantic Web people to follow in twitter. Thanks to @CleverClogs

I’m still trying to make a decision to apply this internship or not.

On the other hand, user contributed data (social media & networks) grows enormously so, we have to spend our time  by eliminate unnecessary data and pick up useful ones. There are two ways to handle this situation; intelligent applications or meaningful data. When we consider development and maintaining costs intelligent applications would be a failure. Giving meaning to data seems more applicable, at this point semantic web plays the role.

By definition;

Semantic web is the extension of the current web which delivers extra information
in well formed machine readable structure.

Web 2.0 is the new way of delivering information with user friendly designs, web services,
social networks, blogs, videos … etc.

Web 3.0 is the evolution of the Web 2.0.

Actually, Web 3.0 is going to be integration of social media with semantic web; adding semantic structure to user contributed data.

So, if you have machine readable data, smart agents can be developed to handle daily tasks you have to do. For example, if you need to buy a book, let it to be done by your agent.  It will seek and find the cheapest book and give an order for you. It is not only for machines, it might change the way we surf the web. For example, semantic web for e-government projects aim to provide all necessary information in one place (website).

However, it might have some issues like security and privacy. Also, semantic structures will result in the rise of agents and new types of spams.

Semantic Web research areas have been increasing and spreading and more companies investing for theses technologies. I hope I can take part in these research projects and develop something useful.

1. Ontology Learning and Metadata Generation (e.g., HLT and ML approaches)
2. Semantics in P2P Computing
3. Semantic Web for e-Business, e-Science, e-Health, e-Culture, e-Government, e-Learning and other application domains
4. Social networks and processes on the Semantic Web

We’ve decided these topics during the meeting before the holiday with my advisers  from both ex and new university. In my opinion these are more promising and familiar to me but I still can change it whatever I like and  any help will be applicable.

Tomorrow night, I will be leaving from Konya and will be at the office Monday morning in İstanbul.