Ranking the whole MEDLINE database according to a large training set using text indexing

Publikationstyp:
Zeitschriftenaufsatz
Metadaten:
Autoren
  • BP Suomela
  • MA Andrade
Autoren-URL
https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=fis-test-1&SrcAuth=WosAPI&KeyUT=WOS:000228242600001&DestLinkType=FullRecord&DestApp=WOS_CPL
DOI
10.1186/1471-2105-6-75
Externe Identifier
  • Clarivate Analytics Document Solution ID: 914QM
  • PubMed Identifier: 15790421
ISSN
1471-2105
Zeitschrift
BMC BIOINFORMATICS
Artikelnummer
ARTN 75
Datum der Veröffentlichung
2005
Status
Published
Titel
Ranking the whole MEDLINE database according to a large training set using text indexing
Sub types
  • Article
Ausgabe der Zeitschrift
6

Datenquelle: Web of Science (Lite)

Andere Metadatenquellen:
Autoren
  • Brian P Suomela
  • Miguel A Andrade
DOI
10.1186/1471-2105-6-75
eISSN
1471-2105
Ausgabe der Veröffentlichung
1
Zeitschrift
BMC Bioinformatics
Sprache
en
Artikelnummer
75
Online publication date
2005
Status
Published online
Herausgeber
Springer Science and Business Media LLC
Herausgeber URL
http://dx.doi.org/10.1186/1471-2105-6-75
Datum der Datenerfassung
2024
Titel
Ranking the whole MEDLINE database according to a large training set using text indexing
Ausgabe der Zeitschrift
6

Datenquelle: Crossref

Abstract
<h4>Background</h4>The MEDLINE database contains over 12 million references to scientific literature, with about 3/4 of recent articles including an abstract of the publication. Retrieval of entries using queries with keywords is useful for human users that need to obtain small selections. However, particular analyses of the literature or database developments may need the complete ranking of all the references in the MEDLINE database as to their relevance to a topic of interest. This report describes a method that does this ranking using the differences in word content between MEDLINE entries related to a topic and the whole of MEDLINE, in a computational time appropriate for an article search query engine.<h4>Results</h4>We tested the capabilities of our system to retrieve MEDLINE references which are relevant to the subject of stem cells. We took advantage of the existing annotation of references with terms from the MeSH hierarchical vocabulary (Medical Subject Headings, developed at the National Library of Medicine). A training set of 81,416 references was constructed by selecting entries annotated with the MeSH term stem cells or some child in its sub tree. Frequencies of all nouns, verbs, and adjectives in the training set were computed and the ratios of word frequencies in the training set to those in the entire MEDLINE were used to score references. Self-consistency of the algorithm, benchmarked with a test set containing the training set and an equal number of references randomly selected from MEDLINE was better using nouns (79%) than adjectives (73%) or verbs (70%). The evaluation of the system with 6,923 references not used for training, containing 204 articles relevant to stem cells according to a human expert, indicated a recall of 65% for a precision of 65%.<h4>Conclusion</h4>This strategy appears to be useful for predicting the relevance of MEDLINE references to a given concept. The method is simple and can be used with any user-defined training set. Choice of the part of speech of the words used for classification has important effects on performance. Lists of words, scripts, and additional information are available from the web address http://www.ogic.ca/projects/ks2004/.
Addresses
  • Ontario Genomics Innovation Centre, Ottawa Health Research Institute, 501 Smyth Rd, Ottawa, Ontario K1H 8L6, Canada. bsuomela@ohri.ca
Autoren
  • Brian P Suomela
  • Miguel A Andrade
DOI
10.1186/1471-2105-6-75
eISSN
1471-2105
Externe Identifier
  • PubMed Identifier: 15790421
  • PubMed Central ID: PMC1274266
Open access
true
ISSN
1471-2105
Zeitschrift
BMC bioinformatics
Schlüsselwörter
  • Stem Cells
  • Humans
  • False Positive Reactions
  • Bayes Theorem
  • Language
  • Computational Biology
  • Algorithms
  • Natural Language Processing
  • Database Management Systems
  • User-Computer Interface
  • Databases, Bibliographic
  • Abstracting and Indexing
  • Vocabulary, Controlled
  • Subject Headings
  • Information Storage and Retrieval
  • MEDLINE
  • Databases, Protein
  • Information Systems
  • Databases as Topic
Sprache
eng
Medium
Electronic
Online publication date
2005
Open access status
Open Access
Paginierung
75
Datum der Veröffentlichung
2005
Status
Published
Publisher licence
CC BY
Datum der Datenerfassung
2005
Titel
Ranking the whole MEDLINE database according to a large training set using text indexing.
Sub types
  • Research Support, Non-U.S. Gov't
  • research-article
  • Journal Article
Ausgabe der Zeitschrift
6

Files

https://bmcbioinformatics.biomedcentral.com/counter/pdf/10.1186/1471-2105-6-75 https://europepmc.org/articles/PMC1274266?pdf=render

Datenquelle: Europe PubMed Central

Abstract
BACKGROUND: The MEDLINE database contains over 12 million references to scientific literature, with about 3/4 of recent articles including an abstract of the publication. Retrieval of entries using queries with keywords is useful for human users that need to obtain small selections. However, particular analyses of the literature or database developments may need the complete ranking of all the references in the MEDLINE database as to their relevance to a topic of interest. This report describes a method that does this ranking using the differences in word content between MEDLINE entries related to a topic and the whole of MEDLINE, in a computational time appropriate for an article search query engine. RESULTS: We tested the capabilities of our system to retrieve MEDLINE references which are relevant to the subject of stem cells. We took advantage of the existing annotation of references with terms from the MeSH hierarchical vocabulary (Medical Subject Headings, developed at the National Library of Medicine). A training set of 81,416 references was constructed by selecting entries annotated with the MeSH term stem cells or some child in its sub tree. Frequencies of all nouns, verbs, and adjectives in the training set were computed and the ratios of word frequencies in the training set to those in the entire MEDLINE were used to score references. Self-consistency of the algorithm, benchmarked with a test set containing the training set and an equal number of references randomly selected from MEDLINE was better using nouns (79%) than adjectives (73%) or verbs (70%). The evaluation of the system with 6,923 references not used for training, containing 204 articles relevant to stem cells according to a human expert, indicated a recall of 65% for a precision of 65%. CONCLUSION: This strategy appears to be useful for predicting the relevance of MEDLINE references to a given concept. The method is simple and can be used with any user-defined training set. Choice of the part of speech of the words used for classification has important effects on performance. Lists of words, scripts, and additional information are available from the web address http://www.ogic.ca/projects/ks2004/.
Date of acceptance
2005
Autoren
  • Brian P Suomela
  • Miguel A Andrade
Autoren-URL
https://www.ncbi.nlm.nih.gov/pubmed/15790421
DOI
10.1186/1471-2105-6-75
eISSN
1471-2105
Externe Identifier
  • PubMed Central ID: PMC1274266
Zeitschrift
BMC Bioinformatics
Schlüsselwörter
  • Abstracting and Indexing
  • Algorithms
  • Bayes Theorem
  • Computational Biology
  • Database Management Systems
  • Databases as Topic
  • Databases, Bibliographic
  • Databases, Protein
  • False Positive Reactions
  • Humans
  • Information Storage and Retrieval
  • Information Systems
  • Language
  • MEDLINE
  • Natural Language Processing
  • Stem Cells
  • Subject Headings
  • User-Computer Interface
  • Vocabulary, Controlled
Sprache
eng
Country
England
Paginierung
75
PII
1471-2105-6-75
Datum der Veröffentlichung
2005
Status
Published online
Datum, an dem der Datensatz öffentlich gemacht wurde
2006
Titel
Ranking the whole MEDLINE database according to a large training set using text indexing.
Sub types
  • Journal Article
  • Research Support, Non-U.S. Gov't
Ausgabe der Zeitschrift
6

Datenquelle: PubMed

Beziehungen:
Eigentum von

Werkzeuge