Completed Research Topics


Temporal entity extraction from historical texts

Behavioural Simulator for Professional Training based on Natural Language Interaction

NCCR Democracy


Answer Extraction

We are developing ways of accessing the content of texts in an intelligent manner, looking for (exact) answers to questions phrased in natural language, or extracting certain types of information automatically from very large volumes of text. More

Automated Style Checking

We conduct research in the area of automated style checking. We focus particularly on the domain-specific stylistic analysis of Swiss legislative texts. More

Controlled Language

We investigate how the ambiguity of natural language can be controlled by restricting its lexicon, syntax and semantics. More

Coreference Resolution

Coreference is the phenomenon of natural languages to have different surface forms to refer to the same entity ("Obama" – "the president" – "he"). We investigate automated detection and resolution of such references. More

Linguistically supported revising and editing – LingURed

We investigate how to prevent typical errors when writing and revising texts (e.g., incorrect word order or sentences without verbs). More

NLP for Quechua

We create Language Technology resources for Cuzco Quechua, an indigenous language spoken by some 1.5 million people in Peru. The resources range from a morphology analyser and spell checker to machine translation and parallel, annotated corpora (treebanks). More

Syntax Analysis and Grammar Construction

We develop methods to syntactically analyse large volumes of text, and create to this end hybrid parsers (partly rule-based, partly statistical), and we develop suitable grammars for various languages (for the time being, Western European languages). More

Automatic Translation of German Train Announcements into Swiss German Sign Language

In the Trainslate (=train+translate) project we are developing a system that automatically translates German train announcements of the Swiss Federal Railways (Schweizerische Bundesbahnen, SBB) into Swiss German Sign Language. The idea for such a system was suggested to us by Deaf signers in Switzerland.


MODERN: Discourse in Statistical Machine Translation

The Institute of Computational Linguistics investigates the integration of cross-sentence features in Statistical Machine Translation (SMT) systems to improve the quality of their translations. This research is motivated by the state-of-the-art Machine Translation (MT) systems which operate at sentence-level and do not propagate information through the sentences.




Sentiment Analysis

  • CTI: Expectation-based sentiment analysis (Jan 2015 - Dec 2015) More
  • CTI: Behavioural Simulator for Professional Training based on Natural Language Interaction (2016–2017) More