A balanced approach to health information evaluation: A vocabulary-based naïve Bayes classifier and readability formulas

Gondy Augusta Leroy, Trudi Miller, Graciela Rosemblat, Allen Browne

Research output: Contribution to journalArticle

21 Citations (Scopus)

Abstract

Since millions seek health information online, it is vital for this information to be comprehensible. Most studies use readability formulas, which ignore vocabulary, and conclude that online health information is too difficult. We developed a vocabularly-based, naïve Bayes classifier to distinguish between three difficulty levels in text. It proved 98% accurate in a 250-document evaluation. We compared our classifier with readability formulas for 90 new documents with different origins and asked representative human evaluators, an expert and a consumer, to judge each document. Average readability grade levels for educational and commercial pages was 10th grade or higher, too difficult according to current literature. In contrast, the classifier showed that 70-90% of these pages were written at an intermediate, appropriate level indicating that vocabulary usage is frequently appropriate in text considered too difficult by readability formula evaluations. The expert considered the pages more difficult for a consumer than the consumer did.

Original languageEnglish (US)
Pages (from-to)1409-1419
Number of pages11
JournalJournal of the American Society for Information Science and Technology
Volume59
Issue number9
DOIs
StatePublished - Jul 2008
Externally publishedYes

Fingerprint

health information
vocabulary
Classifiers
Health
school grade
evaluation
expert
Evaluation
Readability
Health information
Classifier

ASJC Scopus subject areas

  • Information Systems
  • Library and Information Sciences

Cite this

A balanced approach to health information evaluation : A vocabulary-based naïve Bayes classifier and readability formulas. / Leroy, Gondy Augusta; Miller, Trudi; Rosemblat, Graciela; Browne, Allen.

In: Journal of the American Society for Information Science and Technology, Vol. 59, No. 9, 07.2008, p. 1409-1419.

Research output: Contribution to journalArticle

@article{48747bfb7e064e4ba57b0930b2127049,
title = "A balanced approach to health information evaluation: A vocabulary-based na{\"i}ve Bayes classifier and readability formulas",
abstract = "Since millions seek health information online, it is vital for this information to be comprehensible. Most studies use readability formulas, which ignore vocabulary, and conclude that online health information is too difficult. We developed a vocabularly-based, na{\"i}ve Bayes classifier to distinguish between three difficulty levels in text. It proved 98{\%} accurate in a 250-document evaluation. We compared our classifier with readability formulas for 90 new documents with different origins and asked representative human evaluators, an expert and a consumer, to judge each document. Average readability grade levels for educational and commercial pages was 10th grade or higher, too difficult according to current literature. In contrast, the classifier showed that 70-90{\%} of these pages were written at an intermediate, appropriate level indicating that vocabulary usage is frequently appropriate in text considered too difficult by readability formula evaluations. The expert considered the pages more difficult for a consumer than the consumer did.",
author = "Leroy, {Gondy Augusta} and Trudi Miller and Graciela Rosemblat and Allen Browne",
year = "2008",
month = "7",
doi = "10.1002/asi.20837",
language = "English (US)",
volume = "59",
pages = "1409--1419",
journal = "Journal of the Association for Information Science and Technology",
issn = "2330-1635",
publisher = "John Wiley and Sons Ltd",
number = "9",

}

TY - JOUR

T1 - A balanced approach to health information evaluation

T2 - A vocabulary-based naïve Bayes classifier and readability formulas

AU - Leroy, Gondy Augusta

AU - Miller, Trudi

AU - Rosemblat, Graciela

AU - Browne, Allen

PY - 2008/7

Y1 - 2008/7

N2 - Since millions seek health information online, it is vital for this information to be comprehensible. Most studies use readability formulas, which ignore vocabulary, and conclude that online health information is too difficult. We developed a vocabularly-based, naïve Bayes classifier to distinguish between three difficulty levels in text. It proved 98% accurate in a 250-document evaluation. We compared our classifier with readability formulas for 90 new documents with different origins and asked representative human evaluators, an expert and a consumer, to judge each document. Average readability grade levels for educational and commercial pages was 10th grade or higher, too difficult according to current literature. In contrast, the classifier showed that 70-90% of these pages were written at an intermediate, appropriate level indicating that vocabulary usage is frequently appropriate in text considered too difficult by readability formula evaluations. The expert considered the pages more difficult for a consumer than the consumer did.

AB - Since millions seek health information online, it is vital for this information to be comprehensible. Most studies use readability formulas, which ignore vocabulary, and conclude that online health information is too difficult. We developed a vocabularly-based, naïve Bayes classifier to distinguish between three difficulty levels in text. It proved 98% accurate in a 250-document evaluation. We compared our classifier with readability formulas for 90 new documents with different origins and asked representative human evaluators, an expert and a consumer, to judge each document. Average readability grade levels for educational and commercial pages was 10th grade or higher, too difficult according to current literature. In contrast, the classifier showed that 70-90% of these pages were written at an intermediate, appropriate level indicating that vocabulary usage is frequently appropriate in text considered too difficult by readability formula evaluations. The expert considered the pages more difficult for a consumer than the consumer did.

UR - http://www.scopus.com/inward/record.url?scp=47049095650&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=47049095650&partnerID=8YFLogxK

U2 - 10.1002/asi.20837

DO - 10.1002/asi.20837

M3 - Article

AN - SCOPUS:47049095650

VL - 59

SP - 1409

EP - 1419

JO - Journal of the Association for Information Science and Technology

JF - Journal of the Association for Information Science and Technology

SN - 2330-1635

IS - 9

ER -