PhD Researcher on concept recognition and disambiguation in biomedical literature and clinical records. (Erasmus University Medical Center of Rotterdam)
A full-time 4 year PhD position is available as part of a national effort to create the 'Commontology', a structured representation of knowledge contained in unstructured text.
JOB DESCRIPTION: The PhD researcher will investigate, develop, and evaluate natural language processing techniques for the unambiguous and accurate detection of entities in text, using user feedback to automatically improve detection performance. The main application domain will be genomics with its highly ambiguous nomenclature.
PROJECT DESCRIPTION: A lot of scientific knowledge is contained in unstructured text, such as the scientific literature. The first step in extracting this knowledge is identifying the relevant concepts mentioned in the text. Concept recognition in the biomedical domain is extremely difficult due to a wide use of synonyms (several names for the same entity) and homonyms (several entities with the same name). In this project, the PhD student will investigate ways to improve the recognition of concepts by using a wide range of information sources, such as the text around the ambiguous terms, background knowledge about concepts, and background knowledge about the text (e.g. the journal or the author). A key element will be user feedback: users will be able to correct the system, and the system should learn from this feedback.
WHAT WE OFFER: The position will be in in the Biosemantics group (http://biosemantics.org), a multidisciplinary group spanning the Medical Informatics department at the ErasmusMC and the Human Genetics department at the Leiden University Medical Center. This group has a strong international track record in applying text-mining in the biomedical domain, and has an extensive text-mining software and hardware infrastructure.
WHAT WE LOOK FOR: An enthusiastic young researcher interested in natural language processing and machine learning, and willing to gain a background understanding of the biomedical domain.
REQUIREMENTS:
APPOINTMENT, SALARY, LOCATION: The appointment will be at the Erasmus University Medical Center of Rotterdam, and will initially be for 1 year. After successful evaluation, the appointment will be extended by another 3 years, resulting in a dissertation. The gross salary starts at EU 2.435 per month.
The PhD student will work in the Department of Medical Informatics of the Erasmus University Medical Center
INTERESTED? For more information please contact Dr. Martijn Schuemie ( ). Please send your application to Desiree de Jong ( ) by email before April 1 2010.