Dataset Information

A neural classification method for supporting the creation of BioVerbNet.

ABSTRACT:

Background

VerbNet, an extensive computational verb lexicon for English, has proved useful for supporting a wide range of Natural Language Processing tasks requiring information about the behaviour and meaning of verbs. Biomedical text processing and mining could benefit from a similar resource. We take the first step towards the development of BioVerbNet: A VerbNet specifically aimed at describing verbs in the area of biomedicine. Because VerbNet-style classification is extremely time consuming, we start from a small manual classification of biomedical verbs and apply a state-of-the-art neural representation model, specifically developed for class-based optimization, to expand the classification with new verbs, using all the PubMed abstracts and the full articles in the PubMed Central Open Access subset as data.

Results

Direct evaluation of the resulting classification against BioSimVerb (verb similarity judgement data in biomedicine) shows promising results when representation learning is performed using verb class-based contexts. Human validation by linguists and biologists reveals that the automatically expanded classification is highly accurate. Including novel, valid member verbs and classes, our method can be used to facilitate cost-effective development of BioVerbNet.

Conclusion

This work constitutes the first effort on applying a state-of-the-art architecture for neural representation learning to biomedical verb classification. While we discuss future optimization of the method, our promising results suggest that the automatic classification released with this article can be used to readily support application tasks in biomedicine.

SUBMITTER: Chiu B

PROVIDER: S-EPMC6339329 | biostudies-literature | 2019 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A neural classification method for supporting the creation of BioVerbNet.

Chiu Billy B Majewska Olga O Pyysalo Sampo S Wey Laura L Stenius Ulla U Korhonen Anna A Palmer Martha M

Journal of biomedical semantics 20190118 1

<h4>Background</h4>VerbNet, an extensive computational verb lexicon for English, has proved useful for supporting a wide range of Natural Language Processing tasks requiring information about the behaviour and meaning of verbs. Biomedical text processing and mining could benefit from a similar resource. We take the first step towards the development of BioVerbNet: A VerbNet specifically aimed at describing verbs in the area of biomedicine. Because VerbNet-style classification is extremely time c ...[more]

PMID: 30658707

Similar Datasets

Project description:ImportanceClassification criteria are the standardized definitions that are used to enroll uniform cohorts for research studies. They emphasize high specificity and are distinct from diagnostic criteria. No universally recognized classification criteria currently exist for discoid lupus erythematosus (DLE), which has led to problematic heterogeneity in observational and interventional clinical studies across the field.ObjectiveTo create and validate classification criteria for DLE using 12 previously defined candidate criteria items.Design, setting, and participantsFor this diagnostic study, candidate criteria items were prospectively applied by dermatologists and dermatopathologists at clinical visits of patients with DLE or a condition that could be confused for DLE, termed a DLE mimicker, at academic dermatology practices across the United States, Poland, Japan, and South Korea. Data were collected from December 1, 2017, to February 1, 2019, and analyzed from March 1 to September 19, 2019.Main outcomes and measuresClinical features among these 2 groups were calculated and compared with χ2 or Fisher exact tests. Candidate models were identified using best subsets logistic regression analysis. Improvement tests, fit statistics, and discrimination were considered to choose a final model.ResultsNine sites contributed 215 patients, 15 of whom had missing or incomplete data. The final model for DLE classification criteria includes only clinical variables: atrophic scarring (3 points), location in the conchal bowl (2 points), preference for the head and neck (2 points), dyspigmentation (1 point), follicular hyperkeratosis and/or plugging (1 point), and erythematous to violaceous in color (1 point), with an area under the receiving operating characteristic curve of 0.91 (95% CI, 0.87-0.95). A score of at least 5 points yields a sensitivity of 84.1% and a specificity of 75.9% in the classification of DLE, with increasing scores yielding higher specificity.Conclusions and relevanceThese findings provide the initial validation of classification criteria for DLE for use in observational and clinical trials.

Dataset Information

A neural classification method for supporting the creation of BioVerbNet.

Background

Results

Conclusion

Publications

A neural classification method for supporting the creation of BioVerbNet.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets