Sequential Labeling Using Deep-Structured Conditional Random Fields

We develop and present the deep-structured conditional random field (CRF), a multi-layer CRF model in which each higher layer's input observation sequence consists of the previous layer's observation sequence and the resulted frame-level marginal probabilities. Such a structure can closely approximate the long-range state dependency using only linear-chain or zeroth-order CRFs by constructing features on the previous layer's output (belief). Although the final layer is trained to maximize the log-likelihood of the state (label) sequence, each lower layer is optimized by maximizing the frame-level marginal probabilities. In this deep-structured CRF, both parameter estimation and state sequence inference are carried out efficiently layer-by-layer from bottom to top. We evaluate the deep-structured CRF on two natural language processing tasks: search query tagging and advertisement field segmentation. The experimental results demonstrate that the deep-structured CRF achieves word labeling accuracies that are significantly higher than the best results reported on these tasks using the same labeled training set.