About
1. About me
2. About this project
3. About this project's idea
4. About this website
5. Downsides and problems
6. Contact
7. More
About me
My name is Andrei and I've graduated "Politehnica" University of Bucharest's Computer Science College in 2007. My major is Artificial Intelligence and this website is the front-end of my graduation project. For more details about me, just ask.
About this project
In one interpretation, Smartlearned stands for SMART LEarning ARtificial News EDitor. This title tries to describe what this application actually does. Behind every edition of this online publication there is a process of selection. The difference between this website and a regular IT news website is that the role of the human editor is taken by the artificial editor developed by me. Along with the editor itself, other application components make this online publication to manage itself, being completely autonomous (no humans involved).
The artificial editor has to select from all the articles that the application (automatically) grabs from different internet news streams. It does so by analyzing each article and establishing a score that measures the importance of that particular article. Only articles with higher scores make it to the website.
The learning part is related to the way the artificial editor evaluates each article. Its evaluation is based on the popularity of previous articles published on the website (number of user visits). In other words, it learns from its users which subjects are more interesting to them and adapts its editorial policy accordingly.
About this project's idea
In a time when information availability and quantity exceeds more than ever before the humans' comprehending capabilities, systems that perform pre-selections of information pieces such as articles are one solution to keep people efficiently informed regarding their interests. This project specializes in IT-related news, but the engine behind it can be used for any kind of information. This project was my way of tackling the efficiency of the informing process problem, a problem that I believe is grater than it has ever been in the past.
About this website
As I was saying, this website is the front-end of the above-mentioned project.
In some way, it is an indisputable proof of my totally lack of any kind of graphics designing skills :D. Actually, I begged my brother to come up with a simple clean text-oriented design to serve as user interface (he being a 10th grader at the time). Anyway if there is one thing this website didn't intended to do was to be eye-candy.
From the functional point of view, the website presents its current edition's headlines on the main page. From the left menu you can go to a specific category. You can also access the archive (consisting of all the editions) and perform a search within all the articles that where ever published.
Downsides and problems
Regardless of the ideas, concepts, algorithms and working hours put into a product, its degree of utility is after all measured in popularity. Besides that, this particular product, as I've described above, becomes better as its popularity rises (the artificial editor learn from user visits).
The popularity of this website is still low and there are several reasons for that. An inexperienced human editor can at least take some decisions based on his common sense and background, an artificial editor lacks both of these. So as long as this website doesn't have too many visitors, my editor won't be too bright either. But then again, the visitors come in large numbers if an experienced editor chooses some interesting articles for them, and this is a circle hard to break at the beginning.
Maybe even more important is the lack of sympathy the search engines show to this website. Even though is by far my most interested reader, the only one who's read almost all my articles, Google still doesn't want to put this website in top positions as it displays search results. No Google displays, no visitors: everybody knows that. And there's a very simple explanation for that: Google obviously doesn't appreciate the lack of original content on this website, as all these articles appear on other major websites. And there is another issue. For those of you that do not know, one of Google's measures of the importance of a website is the quantity and quality of other websites that refer to it. Well, in my case, there isn't a singe website referring to smartlearned.com. This is obviously my fault because I didn't take the time to perform minimal promoting activities. This is mainly why Google (and other search engines) is so skeptical about smartlearned.com.
Contact
You can contact me at admin@smartlearned.com.
More
The documentation for this project will be available soon and I will also try to make some time to improve this "About" section.