Avoiding Echo-Responses in a Retrieval-Based Conversation System
This addresses a specific issue in conversational AI for improving response quality, but it is incremental as it builds on existing retrieval-based methods.
The paper tackled the echoing problem in retrieval-based conversation systems, where responses overly similar to the context degrade quality, and by using hard negative mining during training, the model reduced echoing and improved Average Precision and Recall@N metrics.
Retrieval-based conversation systems generally tend to highly rank responses that are semantically similar or even identical to the given conversation context. While the system's goal is to find the most appropriate response, rather than the most semantically similar one, this tendency results in low-quality responses. We refer to this challenge as the echoing problem. To mitigate this problem, we utilize a hard negative mining approach at the training stage. The evaluation shows that the resulting model reduces echoing and achieves better results in terms of Average Precision and Recall@N metrics, compared to the models trained without the proposed approach.