Résumé
Semantically analyze patient-generated text from a biomedical perspective is challenging because of the vocabulary gap between patients and health professionals. The medical expertise and vocabulary is well formalized in standards terminologies and ontologies, which enable semantic analysis of expert-generated text; however resources which formalize the vocabulary of health consumers (patients and their family, laypersons in general) remain scarce. The situation is even worse if one is interested in another language than English. In previous studies, we attempted to produce a French preliminary Consumer Health Vocabulary (CHV) by mining the language used within online public forums & Facebook groups about breast cancer. In this work, we show our effort to concretely align the vocabulary produced to standard terminologies and to represent its content (terms & mappings) using semantic web languages such as RDF and SKOS. We used a sample of 173 relations built around 64 expert concepts which have been automatically (89%) or manually (11%) aligned to standard biomedical terminologies, in our case: MeSH, MedDRA and SNOMEDint. The resulting vocabulary, called MuEVo (Multi-Expertise Vocabulary) and the mappings are publicly available in the SIFR BioPortal French biomedical ontology repository.