I disagree also with important parts of the terminology used in that Wikipedia article. Again, it is not scientifically correct to equate a learning process inside the human mind with what computers are programmed to do: crunching numbers (input) and giving back some kind of result (output). The input of your programme depends on a limited number of “experts” who have a “limited” number of ways to look at photos and decide what — in their eyes and in their eyes only- is supposed to be a “good photo”.
How often are you going to update the data fed into the programme, from what sources and based on what decisions? Finally, your programme is and will always remain incredibly limited and totally dependent on new input.
But my main objection is that it will not teach anybody anything about good photography, although that seems to be its primary goal. It is aimed at and can by its very origin only be aimed at quantity, not quality. What I mean is that the overwhelming and infinite number of variables and parameters involved in making decisions about what is and what is not a “good photograph”, makes it impossible to do what you’re trying to do.
Conclusion: It is not the decision of experts in photography (have you thought about who and what is defining who is and who is not an expert in photography?), melted down into numbers and statistics and logarithms and computer code, that will teach or help people to see what is a good photograph and what not. That goal can only be reached by making photos on a regular basis, watching photos on a regular basis, building up solid experiences as a technical photographer and as a photographic artist, and — most certainly- by the interaction between photographers and other artists.
I have nothing against your tool. I have something against what it proclaims to be and can in fact never be. It cannot teach people anything. It can only help them as a tool. It cannot replace the human factor. You might know that and I might know that but all those people out there in the big world of portable computing, do they know that?! I’m afraid most people will just do click-click-click and no longer think about what they are doing or how they are doing it. They will no longer be critical, as something (your application) and someone else (the “experts” involved in your project) will think and decide for them.