LINGUIST List 21.1957
Fri Apr 23 2010
Support: Computational Ling: PhD Student, U. of Copenhagen, Denmark
Editor for this issue: Matthew Lahrman
Computational Linguistics: PhD Student, University of Copenhagen, Denmark
Message 1: Computational Linguistics: PhD Student, University of Copenhagen, Denmark
From: Jürgen Wedekind <jwedekindhum.ku.dk>
Subject: Computational Linguistics: PhD Student, University of Copenhagen, Denmark
E-mail this message to a friend
Institution/Organization: University of Copenhagen
Web Address: http://cst.ku.dk/
Duties: Research,Project Work
Specialty Areas: Computational Linguistics
PhD Scholarship in Question Answering
Within the framework of the ESICT (Experience-oriented Sharing of healthknowledge via Information and Communication Technology) research project, Centrefor Language Technology wishes to appoint a PhD scholar from September 1, 2010or later for a duration of up to three years. Place of employment will be at theFaculty of Humanities, University of Copenhagen.
This PhD will be part of a research team consisting of two additional PhDstudents of the ESICT project and of researchers from the University ofCopenhagen, the Technical University of Denmark and the University MedicalCenter Freiburg. The ESICT project is supported by the Danish Council forStrategic Research.
Project descriptionThe ESICT project aims at providing citizens with an innovative Danishinformation system on health and disease based on information technology,language technology and formalized medical knowledge. The ESICT system willallow users to ask questions in natural language which will be answered the sameway. Such a system addresses demands expected to become increasingly important;citizens must be able to take more responsibility for their own health, e.g. bynavigating available information and acting in accordance with disease relatedinformation provided.
The project is highly cross-disciplinary by using and creating methodologiesrelated to three broad subject areas: information technology, humanities andhealth science. The basic idea is ambitious: to employ relations, descriptionsand concepts from ontologies such as the SNOMED CT (Systematized Nomenclature ofMedicine-Clinical Terms) ontology and extract knowledge from other medical datasources in order to generate answers for the questions formulated by thecitizens. Focusing on two widespread diseases as pilot applications, the projectwill demonstrate feasibility and costs of a large-scale application includingother medical fields.
PhD research projectThe work of the PhD student will consist mainly of research within the field ofquestion answering. The main task will be to investigate and implement thequestion analysis and answer generation components of the system, probably as ahybrid solution where ontological as well as statistical approaches are included.
In a question answering system the information to be presented as an answer isoften not available in a well-worded form, but is rather distributed overvarious resources, and thus the question (i.e. its vocabulary and structuralpattern) becomes the primary basis not only for retrieval of information, butalso for the wording of an answer. Research tasks can include elaboration oftypologies of question/answer patterns and improved methods for automaticidentification of the relevant answer type, e.g. a list, yes/no or a fullsentence. Research in esp. linguistically correct answers is still a relativelyunexplored field and therefore allows the PhD student to contributesubstantially to this research field.
The appointee should be interested in the following topics and havequalifications within more than one of them:- Danish syntax- Natural language parsing and generation- Question answering systems- Knowledge representation- Natural language semantics- Formal logic- Machine learning- Corpus linguistics
Additional information on qualification requirements and the applicationguidelines can be found in the full announcement athttp://www.humanities.ku.dk/research/PhD/Announcements/questionanswering/
Applications Deadline: 14-May-2010
Web Address for Applications: http://www.efond.dk/ansogning/index1.php?fondnr=30&ansoeger_sprog=44
Jürgen Wedekind jwedekindhum.ku.dk