• Sentiment Analysis

​I have been working on a Sentiment Analysis project which our department has carried out in cooperation with ARGUS der Presse (nowadays ARGUS DATA INSIGHTS), a media monitoring company. 

• Computer-Based Media Content Analysis | Framing

In October 2013, I started my PhD about Computer-Based Media Content Analysis which is only one project among ten other groups with their projects in the NCCR democracy research program (3rd phase). This research program focuses on the challenges to democracy in the 21st century and traces them to two major trends: globalization and the growing influence of the media on politics.

The main goal will be to provide access to enriched media data (using Named Entity Recognition, Text Mining, Sentiment Analysis, Relation Mining und several Corpus Linguistics methods) to the other projects and to make the next steps towards a fully automated media content analysis.

Furthermore, we will implement an approach which will be based on data-driven, inductive framing analysis. In this special case we are going to explore further semantically oriented methods coming from the realm of computational linguistics. Especially the coupling of (structured) topic modeling and target-specific, aspect-based sentiment analyis is used to make the shifts in topic and tonality visible. A more profound and updated description of the project and its role and integration in the entire NCCR democracy phase III can be found here on the NCCR democracy website.

Results from an experiment reproducing manual annotation on a large scale corpus can be found here. We used an paragraph-based modeling approach relying on embedded sentences in combination with quantized centroid similarity which allowed for inclusion of knowledge-based lexical resources while still providing the robustness of data-driven methods.

Wueest, Bruno / Schneider, Gerold / Amsler, Michael (2014): Measuring the Public Accountability of New Modes of Governance. In: Proceedings of the ACL 2014 Workshop on Language Technologies and Computational Social Science. Pages 38-43.

Klenner, Manfred / Amsler, Michael / Hollenstein, Nora (2014): Inducing Domain-specific Noun Polarity Guided by Domain-independent Polarity Preferences of Adjectives. In: Proceedings of the 5th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis. Pages 18-23.

Hollenstein, Nora / Amsler, Michael / Bachmann, Martina / Klenner, Manfred (2014): SA-UZH: Verb-based Sentiment Analysis. In: Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014). Pages 503-507.

Klenner, Manfred / Amsler, Michael / Hollenstein, Nora (2014): Verb Polarity Frames: a New Resource and its Application in Target-specific Polarity Classification. In: Proceedings of KONVENS 2014. Pages 106-115.

Klenner, Manfred / Tron, Susanna / Amsler, Michael / Hollenstein, Nora (2014): The Detection and Analysis of Bi-polar Phrases and Polarity Conflicts. In: Proceedings of 11th International Workshop on Natural Language Processing and Cognitive Science, Venice, Italy, 2014

Rocabert, Jofre / Michi Amsler (2015): When do media report on parliamentary monitoring? PADEMIA Workshop ‘Rethinking Representation? The Changing Environment for Parliamentary Democracy’ 26. – 27 March 2015, Institute for Advanced Studies, Vienna.

Amsler, Michi / Schneider Gerold (2015): Data-Driven and Linguistically Motivated Trend Identification. Symposium on the New Frontiers of Automated Content Analysis in the Social Sciences, 1st – 3rd March 2015, Zurich.

Manucci, Luca / Amsler, Michi (2015): The Five Star Movement: a democratic tsunami or another one-man show?. ECPR General Conference 2015,Section “Challenges to Democracy in Southern Europe”, Panel “Rocking the Boat: Changes in Political Participation in Crisis Era Southern Europe”, 26th – 29th August 2015, Montreal.

Klenner, Manfred; Amsler, Michael (2016).  Sentiframes: a resource for verb-centered german sentiment inference. In: Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC 2016), Portorož (Slovenia), 23 May 2016 - 28 May 2016, 1-4.

Amsler, Michael; Wüest, Bruno (2016).  Legitimacy of New Forms of Governance in Public Discourse - An Automated Media Content Analysis Approach Driven by Techniques of Computational Linguistics. In: PolText 2016. International Conference on the Advances in Computational Analysis of Political Text, Dubrovnik, 14 July 2016 - 16 July 2016, 1-7.

Wüest, Bruno / Amsler, Michael / Schneider, Gerold (2017): SIFT – A language technology toolkit to assess the print media coverage of new forms of governance. (Working Paper No. 95). http://www.nccr-democracy.uzh.ch/publications/workingpaper/wp_95

Manucci, Luca / Amsler, Michi (2017): Where the wind blows: Five Star Movement’s populism, direct democracy and ideological flexibility. In: Italian Political Science Review/Rivista Italiana Di Scienza Politica, 1-24. doi:10.1017/ipo.2017.23