Mixed-initiative dialog management for speech-based interaction with graphical user interfaces

Summary

This article describes new ways to implement speech based interaction. The article focuses on general interaction with "everyday" computer. The conversion-and-control "facilitates direct manipulation of widgets by the possibility to employ a command language, which only requires a single command per function."

Pros

  1. The direct command language offers both fine grain control and fast mouse movement.
  2. Can be quicker when finding certain elements within a larger list of different components. An example of this would be for secretary or medical clerical work.

Cons

  1. Speech recognition can take up to 3 times the length of audio input depending on the size of the audio library. This leads to extremely slow performance.
  2. For performance reasons this model is only good for those with disabilities. An able bodied person may find it annoying to learn commands that only offer slow performance at will below known performance with mouse and keyboard.