Fuzzy ontology-based Medical Information Retrieval
Ghada Besbes, Hajer Baazaoui Zghal · 2016
Medical Information Retrieval (IR) aims to extract relevant medical information from the web, patients' records, electronic books, research articles etc. However, users are generally unfamiliar with medical terms and find difficulties expressing their needs. One interesting solution is to integrate fuzzy ontologies in order to achieve semantic interoperability and offer a way to handle vague and imprecise information regarding the medical field. In this paper, our aim is to perform an effective IR by extending its understanding of ambiguous medical concepts. We propose a fuzzy ontology-based medical IR approach which is composed of three components: (1) personalized medical ontology building, (2) contextualized fuzzification of the personalized medical ontology and (3) a query reformulation process based on the resulting fuzzy ontology. A prototype has been implemented allowing experimental evaluation of the proposal. We have studied how the query reformulation has led to a quality results improvement.