Time Masking: Leveraging Temporal Information in Spoken Dialogue Systems
This work addresses a specific bottleneck in dialogue state tracking for spoken dialogue systems, offering an incremental improvement over prior methods.
The paper tackled the problem of tracking dialogue state in spoken dialogue systems by hypothesizing that using wall-clock time differences between turns is crucial for finer-grained control, and demonstrated that their time mask approach outperforms existing distance-based methods on internal and DSTC2 benchmarks.
In a spoken dialogue system, dialogue state tracker (DST) components track the state of the conversation by updating a distribution of values associated with each of the slots being tracked for the current user turn, using the interactions until then. Much of the previous work has relied on modeling the natural order of the conversation, using distance based offsets as an approximation of time. In this work, we hypothesize that leveraging the wall-clock temporal difference between turns is crucial for finer-grained control of dialogue scenarios. We develop a novel approach that applies a {\it time mask}, based on the wall-clock time difference, to the associated slot embeddings and empirically demonstrate that our proposed approach outperforms existing approaches that leverage distance offsets, on both an internal benchmark dataset as well as DSTC2.