Sentiment Analysis

One of our application-oriented research items is Sentiment Analysis and Opinion Mining. The goal here is to identify and annotate evaluations in texts. Our focus lies on the compositional analysis of phrases and sentences. We strive to describe phenomena such as polarity intensification ("very good film") and polarity inversion ("not a glorious film"). In order to support this kind of research, we annotate data and also have created polarity lexicons for English, French and German.

We are the co-initiators of the Interest Group on German Sentiment Analysis (IGGSA). Together with the other group members, we recently released the MLSA corpus, a multi-layered reference corpus for German Sentiment Analysis. The aim of the group is to bundle the research activities for the German language.

Current projects

Completed projects

  • SNF: Bi-directional Sentiment Composition (2009–2012)
  • CTI: Expectation-based Sentiment Analysis (Jan 2015 – Dec 2015): project description (research)
  • CTI: Behavioural Simulator for Professional Training based on Natural Language Interaction (June 2016 – Sept 2017): project description (research)
  • SNF: Sentiment Inference (Dec 2018 - Dec 2022): project description (research)
  • Argus (2013) (cooperation with company)
  • Netbreeze/Microsoft (2014) (cooperation with company)
  • Eurospider (2015) (cooperation with company)
  • LifeLike (2016) (cooperation with company)