This paper introduces a near-linear time sequential algorithm for cons
tructing a sparse neighborhood cover. This implies analogous improveme
nts (from quadratic to near-linear time) for any problem whose solutio
n relies on network decompositions, including small edge cuts in plana
r graphs, approximate shortest paths, and weight- and distance-preserv
ing graph spanners. In particular, an O(log n) approximation to the k-
shortest paths problem on an n-vertex, E-edge graph is obtained that r
uns in (O) over tilde (n + E + k) time.