Introduction to Support vector machines

hi every 开发者_StackOverflow中文版one i m doing POS tagging using SVM tool but i dont understand how to construct feature vector from training corpus.Can someone help me about this matter.

A feature vector is just a series of numbers, each of which representing a measurement or categorical index at a given position, e.g. v1 = (1,0,2,5,5,0,-2,2.9), v2 = (0.2,1,1,8,0,0,4,7), etc.