Our main task is to build the Complex Concept Builder which shall be usable by political scientists to search for specific, social scientific concepts in large corpora.
This is a hard task because a traditional, key-word based search is not sufficient to
detect the complex phenomena political scientists are looking for. These phenomena include
fine grained differences in linguistic expressions of collective identities.
The tools and technology for this advanced search will be developed on a corpus of newspaper
articles from 1990 to 2011 in English, German and French.
For the Complex Concept Builder, the team at the University of Potsdam — together with the IMS Stuttgart — is responsible to implement the preprocessing steps necessary to carry out complex natural-language-processing algorithms including lemmatisation, syntactical parsing and the detection of proper nouns.
The complex tasks carried out on the basis of the above analysis include the interactive, semi-automatical annotation of further material and the analysis of sentiment, coreference, illocution and attribution.
We will also organise a summer school and a workshop in the 3rd year aiming at computational linguists and political scientists. The summer school will present our tools to students and young researchers of the humanities in order to demonstrate the benefits of our newly created tools to them.
