مارسا12

Text Mining, and biological research:

(1) exploring the benefits resulting from improved relations between Text Mining

and biological ontologies,

(2) technical advances in Text Mining and their application to life science research,

and,

(3) success stories of Text Mining solutions with and without ontological support.

The seminar brought together more than 40 internationally renowned researchers

from all domains mentioned beforehand. The ambience of the seminar is best described

with the concept of a prolonged, lively and heated discussion. The discussion

was mainly driven by the divergence of requirements, goals, and expectations between

the Text Mining and the ontology community. On the other side, a number of

talks have pointed out the successful integration of Text Mining solutions into research

in ontological design and the exploitation of ontological resources for successful

Text Mining solutions.

3 Ontologies for Text Mining – Text Mining for ontologies

One particularly important research question in this area is how terminological resources,

such as ontologies, can best support information retrieval (IR) and information

extraction (IE) solutions and vice versa. In theory we can expect that large terminological

resources cover well the domain knowledge and efficiently contribute to one

Dagstuhl seminar proposal „ Ontologies and Text Mining for Life Science“ 3/5

basic information extraction step, i.e. to named entity recognition (NER), in both IR

and IE. In reality, conceptual resources such as ontologies form poor terminological

resources since they have never been designed to serve this purpose