Continuation, turn change, backchannel, interruption, silence: five events a model predicts from 30 seconds of dual-channel audio.

Source: [HackerNoon](https://hackernoon.com/a-framework-for-conversational-system-modeling?source=rss)

Sponsored