Unifed Contextual Query Rewriting

Abstract

Query rewriting (QR) is an important technique for user friction reduction (i.e. recovering ASR error or system error) and contextual carryover (i.e. ellipsis and co-reference) in conversational AI systems. Recently, generation-based QR models have achieved promising results on these two tasks separately. Although these two tasks have many similarities such as they both use the previous dialogue along with the current request as model input, there is no unifed model to solve them jointly. To this end, we propose a unifed contextual query rewriting model that unifes QR for both reducing friction and contextual carryover purpose. Moreover, we involve multiple auxiliary tasks such as trigger prediction and NLU interpretation tasks to boost the performance of the rewrite. We leverage the text-to-text unifed framework which uses independent tasks with weighted loss to account for task importance. Then we propose new unifed multitask learning strategies including a sequential model which outputs one sentence for multi-tasks, and a hybrid model where some tasks are independent and some tasks are sequentially generated. Our experimental results demonstrate the effectiveness of the proposed unifed learning methods.