In her project ‘Computational Rhetorical Analysis: Transfer Learning Methods for Detecting Persuasion in Text’, Amalie sets out to investigate computational modelling and automatic detection of persuasion techniques at the intersection of Natural Language Processing (NLP) and Machine Learning.

Misinformation and propaganda are recognized as major threats to people’s judgement and informed decision. The spread of misinformation relating to the Covid-19 epidemic is just one recent example. It is not only incorrect facts that constitute a threat, but also the language used, which can lead to deception or mislead people.  While rhetoric, the art of persuasion, is an ancient discipline, its computational analysis, regarding persuasion techniques, is still in its infancy. 

With this data science project on computational modelling and automatic detection of persuasion techniques at the intersection of Natural Language Processing (NLP) and Machine Learning, it is expected that detecting and highlighting persuasion techniques enables readers to more critically reading of texts. This could be done by flagging misleading, deceptive, or biased terminology, and thereby reducing the impact the manipulative and disingenuous content could have on the reader.

Amalie is currently employed as a research assistant at the department, and prior to that she worked as a machine learning specialist at the Alexandra Institute for more than three years. She did her bachelor’s in Mathematics-Economics at Aarhus University, and her master’s degree at University of Edinburgh.

On June 20, Amalie attended the opening ceremony of the Danish Data Science Academy during which the 10 PhD fellows – from all over Denmark - were revealed.

About Danish Data Science Academy

The Novo Nordisk Foundation and VILLUM FOUNDATION have awarded a combined grant of DKK 184.3 million to the Danish Data Science Academy (DDSA), which will support and initiate new initiatives for Danish research institutions and tech companies. Primarily within training of experts and interdisciplinary collaboration, two crucial areas for Denmark to realize its potential within data science.

The main objective of the DDSA Fellowship programme is to attract and educate excellent PhD students and postdocs within data science, achieve scientific excellence and impact, while simultaneously extending the application of data science in relevant scientific domains. All DDSA scholarships are fixed individual awards to the most promising candidates. Funding young researchers is instrumental to obtaining one of the overarching goals of the DDSA: to increase volume and quality of data science research. Furthermore, fellowship holders will be key participants in the efforts to build a data science community across disciplines. More information available at https://ddsa.dk/