Dynamically Context-Sensitive Time-Decay Attention for Dialogue Modeling
This work addresses a specific bottleneck in dialogue systems by enhancing contextual understanding with time-aware attention, though it is incremental in nature.
The paper tackled the problem of incorporating temporal information in dialogue modeling for spoken language understanding by proposing a dynamically context-sensitive time-decay attention mechanism, which improved the state-of-the-art model on the DSTC4 dataset.
Spoken language understanding (SLU) is an essential component in conversational systems. Considering that contexts provide informative cues for better understanding, history can be leveraged for contextual SLU. However, most prior work only paid attention to the related content in history utterances and ignored the temporal information. In dialogues, it is intuitive that the most recent utterances are more important than the least recent ones, and time-aware attention should be in a decaying manner. Therefore, this paper allows the model to automatically learn a time-decay attention function where the attentional weights can be dynamically decided based on the content of each role's contexts, which effectively integrates both content-aware and time-aware perspectives and demonstrates remarkable flexibility to complex dialogue contexts. The experiments on the benchmark Dialogue State Tracking Challenge (DSTC4) dataset show that the proposed dynamically context-sensitive time-decay attention mechanisms significantly improve the state-of-the-art model for contextual understanding performance.