Comparing paths ?

This question is related to the topic "Mouse drawn symbols" somewhere down in the forum. I have come to a point where I can describe a path (the symbol that you draw with the mouse), with a quite satisfying number of details. The problem now comes from comparing an array of data describing the input symbol, and the model symbol. I am using an array of structures, but the problem is that the number of elements in my input array can be different from my model array. So do I simply ignore the missing elements, or is there a nicer way to do the comparison ?? I heard the buzzwords Bayesian Classifiers, or Hidden Markov Models, but the tutorials are inexisting, or made for people who seem to already know the basics about those topics, which I don''t. Any ideas, links ??