We present a unifying approach to the study of entropies in mathematics, such as measure entropy, various forms of topological entropy, several notions of algebraic entropy, and two forms of set-theoretic entropy. We take into account only discrete dynamical systems, that is, pairs (X,φ), where X is the underlying space (e.g., a probability space, a compact topological space, a group, a set) and φ:X→X is a transformation of X (e.g., a measure preserving transformation, a continuous selfmap, a group homomorphism, a selfmap). We see entropies as functions h:X→R+, associating to each flow (X,φ) of a category C either a non-negative real number or ∞. First, we introduce the notion of semigroup entropy hS:S→R+, which is a numerical invariant at...