Other

Multi-System Approach to Speech Understanding

Raj Reddy, NASA Technical Reports Server (NTRS), 1993

Read on October 17, 2026

What it is: A retrospective by Raj Reddy of Carnegie Mellon University on the 1970s speech understanding program. It recalls the 1971 performance goals and reports the results of the Harpy system.
How much weight it carries: Written by a participant in the program, so it describes goals and results from inside the project. It reports semantic error on a limited document retrieval task, not word error rate, and it is not a controlled measurement study.

Facts taken from this source

A group of scientists recommended in 1971 a five-year program whose target was a system that accepts connected speech from many speakers on a 1000-word task vocabulary.
“In 1971, a group of scientists recommended the initiation of a five-year research program ... The system was required to accept connected speech from many speakers based on a 1000 word vocabulary task-oriented grammar” (Introduction)
Carnegie Mellon's Harpy system recognized speech from male and female speakers on a 1,011-word document retrieval task, with a semantic error of 5 percent.
“It recognizes speech from male and female speakers using a 1011-word-vocabulary document retrieval task. Semantic error is 5%” (Results paragraph)