The Department of Biomedical Informatics (D2IM) of Rouen university Hospital has been historically working on manual and automatic indexing of Health resources on the Internet. For more than 20 years, the team has been focusing on using controlled vocabularies (such as terminologies) to represent more diversified contents (bibliographic citations, patient records…). Expertise in terms of knowledge engineering is a major asset since today's data exploitation tools rely mainly on knowledge databases (inference, reasoning, etc.).
Specifically, the D2IM develops and maintains the Health Terminology/Ontology Portal (HeTOP) that contains more than 70 terminologies/ontologies in several languages (over 360,000 health concepts in French). These resources are mapped to create a semantic "super-network" storing knowledge and intelligence.
In addition, the D2IM provides a text annotation tool called ECMT (Extracteur de Concepts Multi-Terminologique). It consists in a set of programs that aims at identifying relevant terminological concepts in texts, especially in medical texts.
The D2IM is also specialized in Information Retrieval and, de facto, in search engines based on controlled vocabularies. For this reason, in addition to the development of a search engine dedicated to the CISMeF catalog and to the LiSSa bibliographic database , the team has conceived a link generator called InfoRoute (infobutton) to various databases (such as PubMed, HAS, BDSP, etc.). It takes advantage of the maximum capabilities of each targeted search engine in terms of dedicated grammars and semantics, to improve the precision and recall of queries formulated in natural language. These tools, based on Java 8, are dealing with a large amount of data and therefore are built on a high performance IMDG (In-Memory Data Grid) empowered by parallel algorithms. This is achieved thanks to a cluster of several servers that are capable of processing such data (56 to 144 cores with 224Go to 1To of RAM).
For some tasks, D2IM will work with the DUMG of Rouen (University Department of General Medicine).