Indexing grey multilingual literature in General Practice: Family Medicine in the Era of Semantic Web.

Jamoulle, M., Pizzanelli, M., Cardillo, E., Ittoo, A., Vander Stichele, R., Resnick, M., Grosjean, J., Darmoni, S., & Vanmeerbeek, M. (December 2018). Indexing grey multilingual literature in General Practice: Family Medicine in the Era of Semantic Web. Grey Journal, Spring 2018 (1).

https://hdl.handle.net/2268/229264

[en] Problem/Goal: Sharing the results of research with General Practitioners (GPs) is crucial for the survival of the discipline of General Practice / Family Medicine (GP/FM). The production of abstracts in GP/FM probably exceeds 15,000 per year worldwide. Each abstract often represents two years of work for its authors and is expressed in local languages. Only 45% of them are published in indexed medical journals. Usual indexing systems like MeSH are not multilingual nor adapted to the particular field of GP/FM. Consequently, these abstracts are lacking bibliographic control and more than half of the research presented by GPs at congresses is lost. Considering the absence of appropriate domain-specific terminologies or classification systems, we propose a new multilingual indexing system. The existing International Classification of Primary Care (ICPC) is currently used for clinical purposes and has now been expanded with a taxonomy related to contextual aspects (called Q-Codes) such as education, research, practice organization, ethics or policy in GP/FM, currently not captured. The set is proposed under the name Core Content Classification in General Practice (3CGP). The aim is to facilitate indexing of GP/FM specific scientific work and to improve performance in information storage and retrieval for research purposes in this field. Research Method/Procedure: Using qualitative analysis, a corpus of 1,702 abstracts from six GP/FM- related European congresses was analyzed to identify main themes discussed by GPs (e.g., continuity, accessibility or medical ethics), handled in a domain-specific taxonomy called Q-Codes and translated in 8 languages. In addition, a methodology for building a lightweight ontology (in OWL-2) was applied to Q-Codes, adding object and datatype properties to the hierarchical relations, including mapping to the MeSH thesaurus, Babelnet (www.babelnet.org) and Dbpedia. Finally, the Q-Codes in 8 languages have been integrated in a healthcare terminology service (www.hetop.eu/q) with a companion website (http://3cgp.docpatient.net). Results of the Research: The creation and the on-line publication of this multilingual terminological resource (see http://3cgp.docpatient.net), for indexing abstracts and for facilitating Medline searches, could reduce loss of knowledge in the domain. The database is available in 12 human languages and in Web Ontology Language (OWL) for computer use. In addition, through better indexing of the grey literature (online database, congress abstracts, master’s and doctoral thesis) of which we show the first results, we hope to enhance the accessibility of research in GP/FM domain and promote the emergence of networks of researchers. Indication of costs related to the project: This project has not been funded. 3CGP is placed under Attribution-Non-Commercial-Share-Alike 4.0 International (CC BY-NC-SA 4.0). ICPC is copyrighted by WONCA. Keywords General practice, Terminology, Electronic publishing, Repository, Grey Literature.