The p-norm is a generalization of the Euclidean norm in a vector space, defined as the p-th root of the sum of the absolute values of the vector's components raised to the power of p. It provides a flexible way to measure vector magnitude, with special cases including the Manhattan norm (p=1), Euclidean norm (p=2), and the maximum norm (p=∞).