Software testing and the naturally occurring data assumption in natural language processing

Authors:
K. Bretonnel Cohen;William A. Baumgartner, Jr.;Lawrence Hunter
Affiliations:
The MITRE Corporation;University of Colorado School of Medicine;University of Colorado School of Medicine
Venue:
SETQA-NLP '08 Software Engineering, Testing, and Quality Assurance for Natural Language Processing
Year:
2008

Citing 7
Cited 3

The craft of software testing: subsystem testing including object-based and object-oriented testing

The craft of software testing: subsystem testing including object-based and object-oriented testing
Managing The Testing Process

Managing The Testing Process
Art of Software Testing

Art of Software Testing
Testing Computer Software, Second Edition

Testing Computer Software, Second Edition
The LinGO Redwoods treebank motivation and preliminary applications

COLING '02 Proceedings of the 19th international conference on Computational linguistics - Volume 2
Manual curation is not sufficient for annotation of genomic databases

Bioinformatics
Validation and regression testing for a cross-linguisic grammar resource

DeepLP '07 Proceedings of the Workshop on Deep Linguistic Processing

Designing testsuites for grammar-based systems in applications

GEAF '08 Proceedings of the Workshop on Grammar Engineering Across Frameworks
Context inducing nouns

KRAQ '08 Coling 2008: Proceedings of the workshop on Knowledge and Reasoning for Answering Questions
Benchmarking for syntax-based sentential inference

COLING '10 Proceedings of the 23rd International Conference on Computational Linguistics: Posters

Quantified Score

Hi-index	0.00

Visualization

Abstract

It is a widely accepted belief in natural language processing research that naturally occurring data is the best (and perhaps the only appropriate) data for testing text mining systems. This paper compares code coverage using a suite of functional tests and using a large corpus and finds that higher class, line, and branch coverage is achieved with structured tests than with even a very large corpus.