With those replies, I double that he will come back

I agree on the fact that you should not start from studying the code of a full game (too confusion and too mush copy/paste).
I don't think that a risk type game is that mush difficult (unless you want to do networking and if you stick with a "dumb" AI).
You don't need fancy graphics or even animations. So a Graphics2D / Swing application is good enough.
So I was thinking of creating a risk type game ,I have a couple of ideeas on how I could achive that but nothing concrete on how I could
implement my ideeas.
I think it is the best starting point. You should give us your ideas on the gameplay/design and how you think it could be done. Then you will get feedbacks and advices on them.