Work
on WP1 has been completed and deliverable D1.1:
State of the Art in Cross-Lingual Information Access for
Medical Information has been submitted to the commission.
It provides a survey of relevant areas of natural language
processing and cross-lingual (medical) information management.
It serves both to ensure that the project draws maximal
benefit from previous work and available resources, thus
avoiding duplication of work, and as a medium for project
partners to share their expertise and to arrive at a common
understanding of the nature and complexity of the tasks
involved. Furthermore, by making the report publicly accessible,
we hope that it may serve as a resource for other researchers
interested in the area. We identified four major topics
and structured the document accordingly.
Concept-based Information Access depends on the
identification of meaningful items and patterns in text
by extracting relevant terminology and semantic relations
between terms. A section on text summarization deals with
the related topic of generating multilingual, multi-document
summaries.
Cross-Lingual Information Retrieval The need to transfer
information between languages adds to the complexity of
information retrieval. The chapter discusses approaches
to this task that rely on a variety of resources, as well
as ways of producing those resources.
Word Sense Disambiguation Determining which sense
a given occurrence of a word has, is an important enabling
task for concept-based, cross-lingual information access.
The chapter discusses various methods and problematic issues
involved.
Medical Knowledge and Terminology A discussion of
the ways in which information is structured and used in
the medical professions, as well as a compilation of available
resources, such as thesauri, nomenclatures and dictionaries.
|
|