Conversational question answering (CQA) is a novel QA task that requires understanding of dialogue context. Different from traditional single-turn machine reading comprehension (MRC) tasks, CQA includes passage comprehension, coreference resolution, and contextual understanding. In this talk, I will introduce an innovative contextualized attention-based deep neural network, SDNet, to fuse context into traditional MRC models. The model leverages both inter-attention and self-attention to comprehend conversation context and extract relevant information from passage. Furthermore, I will demonstrate a novel method to integrate the latest BERT contextual model. Empirical results show the effectiveness of the model, which sets the new state of the art result in CoQA leaderboard, outperforming the previous best model by 1.6% F1. The ensemble model further improves the result by 2.7% in F1.
Chenguang Zhu is a researcher in Microsoft Speech and Dialogue Research Group. His research interest is in natural language processing, focusing on machine reading comprehension, dialogue, and textual embedding. He has achieved top place in several famous NLP competitions, e.g. SQuAD v1.0, ARC, and most recently CoQA. He holds PhD in Computer Science from Stanford University.