Responding to Semantically Ill-Formed Input,
Abstract
One cause of failure language interfaces is semantic overshoot; this is reflected in input sentences which do not correspond to any semantic pattern in the system, We describe a system which provides helpful feedback in such cases by identifying the semantically closest inputs which the system would be able to understand. We propose to provide the user with sentences which are semantically close to the original input (in a sense to be defined below) and are acceptable inputs to the system. Such feedback may occasionally be confusing, but we expect that more often it will be helpful in showing the system's capabilities and suggesting possible rephrasings.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1987
- Accession Number
- ADA188638
Entities
People
- Ping Peng
- Ralph David Grishman
Organizations
- New York University