In this paper, we propose a novel Chinese word segmentation method which leverages the huge deposit of Web documents and search technology. It simultaneously solves ambiguous phrase boundary resolution and unknown word identification problems. Evaluations prove its effectiveness.
