The SJTU System for Dialog State Tracking Challenge 2

Dialog state tracking challenge provides a common testbed for state tracking algorithms. This paper describes the SJTU system submitted to the second Dialogue State Tracking Challenge in detail. In the system, a statistical semantic parser is used to generate refined semantic hypotheses. A large number of features are then derived based on the semantic hypotheses and the dialogue log information. The final tracker is a combination of a rulebased model, a maximum entropy and a deep neural network model. The SJTU system significantly outperformed all the baselines and showed competitive performance in DSTC 2.