Loading…

Benchmarking Text Understanding Systems to Human Performance: An Exploration

This report explores the use of natural language in intelligent computer systems specifically with regard to text understanding systems. The goal of the research was to benchmark selected text understanding systems to human performance in reading comprehension. To this end a reading comprehension te...

Full description

Saved in:
Bibliographic Details
Main Authors: Butler, Frances A, Baker, Eva L, Falk, Tine, Herl, Howard, Jang, Younghee, Mutch, Patricia
Format: Report
Language:English
Subjects:
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This report explores the use of natural language in intelligent computer systems specifically with regard to text understanding systems. The goal of the research was to benchmark selected text understanding systems to human performance in reading comprehension. To this end a reading comprehension test was compiled which included the texts and questions from six intelligent computer systems. This test along with a criterion reading measure was administered to the subjects in the study. The computer systems' performances were then benchmarked to a scale of human performance on reading comprehension as part of the development of the Artificial Intelligence Measurement System (AIMS).