Data Distillation for Controlling Specificity in Dialogue Generation
This work addresses the need for conversational agents to adapt specificity based on context, though it is incremental as it builds on existing neural and reinforcement learning techniques.
The paper tackles the problem of controlling specificity in dialogue generation by proposing a data distillation method that trains multiple models with varying specificity levels and uses reinforcement learning to select the appropriate model for each context, resulting in more interesting and higher-quality responses compared to a baseline model.
People speak at different levels of specificity in different situations. Depending on their knowledge, interlocutors, mood, etc.} A conversational agent should have this ability and know when to be specific and when to be general. We propose an approach that gives a neural network--based conversational agent this ability. Our approach involves alternating between \emph{data distillation} and model training : removing training examples that are closest to the responses most commonly produced by the model trained from the last round and then retrain the model on the remaining dataset. Dialogue generation models trained with different degrees of data distillation manifest different levels of specificity. We then train a reinforcement learning system for selecting among this pool of generation models, to choose the best level of specificity for a given input. Compared to the original generative model trained without distillation, the proposed system is capable of generating more interesting and higher-quality responses, in addition to appropriately adjusting specificity depending on the context. Our research constitutes a specific case of a broader approach involving training multiple subsystems from a single dataset distinguished by differences in a specific property one wishes to model. We show that from such a set of subsystems, one can use reinforcement learning to build a system that tailors its output to different input contexts at test time.