The SemanticMining Project
Hans Åhlfeldt


The general objective of a Network of Excellence is to bridge gaps in the European research infrastructure and to facilitate cross-fertilisation between scientific disciplines. Traditionally academic departments in the domain of medical and bioinformatics have their roots either in computer science, system engineering (including a variety of engineering disciplines) or in a medical or clinical context. The SemanticMining network is therefore composed of partners from these scientific areas. An important aspect is the merging of medical or clinical informatics and bioinformatics. Another bridging activity addressed is knowledge-transfer between academia and organisations in the health and welfare sector, such as classification centres, standardisation bodies and public health care organisations.

The research activities in SemanticMining is focused around seven areas: principles in ontology engineering, evaluation of SNOMED CT, impact of ontologies on health statistics, concept systems in laboratory medicine, the construction of a multi-lingual medical dictionary, text mining and information retrieval in bioinformatics, and the concept-based electronic health record. Researchers in the network play an influential role in the process of harmonisation and further development of terminology systems. Examples of areas of interaction are the Gene Ontology, the Foundational Model of Anatomy, and SNOMED CT. Part of the network objectives is also an active interaction with standardisation bodies such as CEN TC251, IMIA and W3C.

An important part of the network activities are knowledge-transfer and spreading of excellence between the different partners in the project. We are therefore organising a series of workshops and tutorials in relevant areas. A major event was the Summer School, held at the lake Balaton in Hungary in July. During this one week session, workshops were organised in the areas of ontology engineering, the semantic web and health statistics. The richness of the network was utilised in so far that different perspectives on ontology engineering was presented and discusses - from discussions on “ontology” and “concept” in a philosophical and theoretical sense to ontology engineering more as a software developing activity. The workshop on health statistics brought the well known difficulties with comparability and reliability of health statistics based on traditional use of classifications such as ICD10 into focus. The underlying question addressed was “What can ontologies do for health statistics?”

In the presentation I will give a short status report from the SemanticMining project with examples of ongoing research activities and highlights from the Summer School, ending with a preliminary presentation of planned activities for the coming year.