Abstract
Scientists are challenged by an ever-increasing number of scientific papers which makes it hard for them to keep track of the relevant literature. A promising solution of this problem is to provide scientists with computational support accessing and evaluating scientific results. A well-known approach is to use natural language processing (NLP) together with the techniques of text mining and summarizing. In this thesis we suggest another approach: the results of scientific papers are summarized by its authors in a formal language, and these formal summaries are added to the papers. In this way we make the results of scientific papers readable and Ð to some degree Ð understandable by computers. As formal language we use Attempto Controlled English (ACE) that is a controlled natural language [7]. Controlled natural languages combine the natural look and feel of natural languages with the processability of formal languages. This makes them to prime candidates for the formal representation of scientific results. Since they are easy to learn and understand, a researcher has not to spend a lot of time learning a complicated formal language. To demonstrate the practicality of our approach, we build an ontology for protein interactions in ACE and show how scientific results can be expressed using this ontology.

URL:
http://rewerse.net/publications/rewerse-publications.html#REWERSE-TR-2006-01

BibTeX:

@techreport{REWERSE-TR-2006-01,
	author = {Tobias Kuhn},
	title = {Expressing Knowledge about Protein Interactions in Attempto Controlled English},
	institution = {Institute for Informatics, University of Munich},
	year = {2006},
	type = {{research report, REWERSE-TR-2006-01}},
	number = {REWERSE-TR-2006-01},
	note = {Master Thesis, Department of Informatics, University of Zurich, January 2006},
	url = {http://rewerse.net/publications/rewerse-publications.html#REWERSE-TR-2006-01}
}