Abstract: A system and method for remotely grouping contents of an action history stack running on a computer system. An action is performed in a first application which causes a beginning marker to be placed on the action history stack. One or more actions are performed in the first application. Each action performed is also stored on the action history stack. Finally an action is performed which causes an end marker to be placed on the action history stack. When some actions need to be undone, a first action is caused to be undone. If the undone action indicates an end marker, then more actions are caused to be undone until an undone action indicates a beginning marker. This method is useful for any application which uses an action history stack, and where access to the action history stack is limited. In one embodiment, the application is a word processing program, and the steps of the invention are performed by a voice recognition system.

Abstract: In the speech recognition system disclosed herein, the Viterbi decoding of an acoustic recognition network is augmented by implementing additional data structures for each arc in the network which represent, for each arc, the best path cost of reaching the last point on that arc and an arc in time which represents, for the best path, the time of leaving the previous arc. These additional data structures enable a trace back procedure which identifies not only the presumably optimal path but also alternate paths having good scores.