Research

Research

In its research and teaching, the Department of Government primarily focuses on comparative and Austrian politics. Its research is concerned with political behaviour, political actors, such as political parties and politicians, political institutions, the processes governed by these institutions, as well as their outcomes. It includes work on political participation, voting behaviour, parties and party competition, coalition politics and Austrian politics in general and is mostly based on rationalist and behavioural approaches.

Our goal is to conduct high-level, internationally competitive research in the area of Comparative Politics with the collaboration of international project partners and research networks. At the Faculty of Social Sciences the department is mainly engaged in the key research area ''Political Competition and Communication: Democratic Representation in Changing Societies'.

The department’s approach places it in the discipline’s empirical-analytical core and is mostly based on quantitative social science methods. To map empirical phenomena accurately, researcher in the department focus on the continuous development of survey design, as well as on the analysis of empirical data by applying the best suited statistical model. The department aims to achieve the best work on Austrian politics and to make important contributions to the international academic literature on Comparative Government and Politics.

An overview of current publications and activities at the department can be found below and on the personal websites of our team.

Publications

More than Bags of Words: Sentiment Analysis with Word Embeddings

Author(s)
Elena Rudkowsky, Martin Haselmayer, Matthias Wastian, Marcelo Jenny, Stefan Emrich, Michael Sedlmair
Abstract

Moving beyond the dominant bag-of-words approach to sentiment analysis we introduce an alternative procedure based on distributed word embeddings. The strength of word embeddings is the ability to capture similarities in word meaning. We use word embeddings as part of a supervised machine learning procedure which estimates levels of negativity in parliamentary speeches. The procedure's accuracy is evaluated with crowdcoded training sentences; its external validity through a study of patterns of negativity in Austrian parliamentary speeches. The results show the potential of the word embeddings approach for sentiment analysis in the social sciences.

Organisation(s)
Research Group Visualization and Data Analysis, Department of Government
External organisation(s)
Technische Universität Wien, Leopold-Franzens-Universität Innsbruck, dwh GmbH, Jacobs Universität Bremen
Journal
Communication Methods & Measures
Volume
12
Pages
140-157
No. of pages
18
ISSN
1931-2458
DOI
https://doi.org/10.1080/19312458.2018.1455817
Publication date
2018
Peer reviewed
Yes
Austrian Fields of Science 2012
508007 Communication science, 506014 Comparative politics
Keywords
Portal url
https://ucrisportal.univie.ac.at/en/publications/808bdf8a-5c70-4b0d-abaf-5d87b95cfeb5