Jena Tutorials

The HP's Jena Framework has become very popular lately in the Semantic Web development world . I started to notice that many of this blog's visitors came here looking for sample code about it. This is why I thought that I could help them out by providing few links where tutorials can be found.

I have noticed that many of the visitors come here seeking some tutorial or getting-started code samples about Jena , besides the code placed in its documentation. Therefore, I decided to look around the Internet and see if I can find some people who already spent some time writing tutorials about Jena . Fortunately, I found some code, which, in my opinion, could very useful if you happen to be someone starting off writing Semantic Web applications. So, here is a small list of links where Jena tutorials can be found:

• Tutorial: Jena Semantic Web Framework
• Playing with the Jena Semantic Web Framework
• Using Jena to Construct FOAF Data
• Jena Rules Examples

I agree though, that some of those tutorials may be a few years old, but, in fact, they might serve you well in seeing some kick-off code where the Jena documentation lacks it. For example, in one of those tutorials, I saw a  working set of Jena rules - whereas its documentation was not rich on code about it. 

In case you are wondering what Jena is (which I doubt), I wrote an article describing this Semantic Web Framework a couple of months ago. 

Dear readers, have you ever worked with Jena? What is your personal experience about this framework? Would you favor some other Semantic Web framework instead? Why or why not?

Facebook Questions - Another Search Frontline

Facebook is testing the new Questions  feature - a search engine that finds relevant people to answer other people's questions. Interesting, the word "relevant" here plays great role - Facebook must be pulling data and make conclusions based on that data. "Where from ?" - One may ask. Well, the countless "like"s people do, provide that data, which is now semantically annotated. But behind the scenes, this application concerns several rivals, including Google and Twitter.

The People Search Engine

There are trials to revolutionize the way people search the today. In a world, dominated by authoritative, fast and amazingly complex algorithm for textual search, it seems there is not much to be improved. And for now, people are happy - they just type in what they are interested in, and Google (by saying Google, I also count Yahoo!, Bing and others in)  and: BANG! - it appears in the first 4 results displayed. Although today's search engines subtly show their power by answering well even if the user misspells the word or moreover, categorizing it like Posts, Tweets, Images etc. Google nowadays is even faster in updating its search index, reaching the Real Time Web Informer status.


But however, there are situations in which textual searching can't do much help. Queries like: "Hey Google, what are the songs that have reached No. 1 in UK's Chart in the last 10 years ?" or ... "Bing buddy, I am looking for a movie tonight. Basically I want comedies, like 'American Pie' or 'Dumb and Dumber', do you happen to have some recommendations for me?" etc. etc. (Examples for such situations can be quite a few, I will not go any further). Well, some of these queries might have to wait for the Semantic Search Engine to be built, but the interesting thing is there is a new trend : to build Social (People) Engines, which will not scan text, but people and their habits. These People Engines will try to discover the right person to answer human-only interpretable questions like those above. But to achieve that, these search engines need to have additional metadata about each person : her skills, interests, friends etc. Having that in mind, one can easily conclude that  building such network can be a challenging and expensive task to do - unless... unless it is built by itself - like in the example of Facebook. Facebook is lucky for having so connected network with correctly filled information such as people's names, age, photos, interests, friends etc. In my opinion, Facebook is now preparing to take advantage of that metadata it has: to build internal network for answering questions - people will answer each other's questions on any topic - Facebook will only be the platform to find the "relevant" people to answer them.


Facebook Questions Application


This application is exactly that: means for people to use "Social Search" instead of "Textual Search". A smart move, I say, because engaging real  people in answering topic-specific questions (for free!) is the currently best way to get around the technological gap that prevents us from building software agents to answer the questions for us. Whoever came with the idea of building this application, must have studied people behavior and conclude that people would react on such questions, if they feel they are concerned on some point with the topic of the question - be it their profession or simply a good or bad experience of some product. We will still need to wait to see how this invention of Facebook will impact the Tweetosphere and the ordinary search.

Google's Social Search Engine Efforts


Surprise, surprise, but Facebook did not invent this whole People-answering-questions thing. Google has been spending time on this field quite a while, resulting in an experimental application in Google Labs, which returns 20 % of the results from your Google queries as answers based on your social graph within Google and in inquiry of Aardvark (the closest relative of Facebook Questions app). Aardvark gathers each person's interests they type in and parses the question text, matches entities with people's interests and finds people that might be able to answer the questions. I have been looking into this application for a while and indeed it has proven itself to be very useful: most of the times it did find a person to be able to answer my question and ... what I really like about is that it integrates with the IMs : be it MSN Messenger, GTalk or Skype ... Pretty cool.
Therefore, I think that these kind of engines do have bright future, no matter what vendor creates them.

Readers, what do you think ? Would you use some of these services as your secondary search engines ? Do you think they will one day integrate with the textual search engines ?