A Response Retrieval Approach for Dialogue Using a Multi-Attentive Transformer
This work provides an incremental improvement for dialogue systems, specifically for shopping assistants, by enhancing response retrieval accuracy.
The authors developed a multi-attentive transformer model for response retrieval in simulated shopping assistant dialogues. Their model achieved the second-best scores across all retrieval metrics on the SIMMC Fashion Dataset.
This paper presents our work for the ninth edition of the Dialogue System Technology Challenge (DSTC9). Our solution addresses the track number four: Simulated Interactive MultiModal Conversations. The task consists in providing an algorithm able to simulate a shopping assistant that supports the user with his/her requests. We address the task of response retrieval, that is the task of retrieving the most appropriate agent response from a pool of response candidates. Our approach makes use of a neural architecture based on transformer with a multi-attentive structure that conditions the response of the agent on the request made by the user and on the product the user is referring to. Final experiments on the SIMMC Fashion Dataset show that our approach achieves the second best scores on all the retrieval metrics defined by the organizers. The source code is available at https://github.com/D2KLab/dstc9-SIMMC.