Practically anyone can provide "medical" information on the web. Such information can be misleading and even life-threatening if it is trustingly taken up by patients and/or their kins. On a daily basis, medical associations and agencies around Europe are striving to manually label hundreds of new medical web sites of ambiguous quality.

The multi-national MedIEQ project, co-funded by the European Commission’s Public Health programme, commenced its ambitious 3-year research effort on January 1st 2006. It is set to pave the way towards automatic quality labelling of medical web sites.

The project consortium comprises partners in the fields of medical quality labelling, web crawling and spidering, multilingual information extraction and semantic resources from six European countries. MedIEQ will deliver web-crawling tools to locate medical web sites in seven European languages (Spanish, Catalan, German, English, Greek, Czech, and Finnish) in order to verify their content against a set of machine readable quality criteria.

MedIEQ is the continuation of previous EU projects in the area of medical quality labelling (MedCIRCLE, MedCERTAIN, WRAPIN) and quality labelling standards (QUATRO). The latter has built a common platform to enable all types of labelling systems to be machine-readable.

Machine-readable labels will allow search engines to display logos next to search results, for example, so that end users can see immediately which websites have been reviewed by labelling authorities.

Two applications are foreseen in the context of MedIEQ: a) constant monitoring of already labelled medical web sites comparing newly extracted information from the web site pages against existing labelling data stored in the accreditation label, and b) automatic identification of new medical web sites in specific thematic areas, their description against machine-readable criteria, the filtering of some of the identified sites based on their description, and their organisation into web directories in order to facilitate access by health information seekers.

K.V. Chandrinos, Managing director i-sieve technologies ltd

K.V. Chandrinos is the managing director of i-sieve technologies ltd. a technical partner in medIEQ. Before leading i-sieve, a research spin-off on filtering applications from NCSR “Demokritos”, Chandrinos worked many years in research on web content analysis and classification.