Ana Barić

Ana Barić is a PhD student at TakeLab FER, focused on NLP. Her research interests span from exploring model uncertainty in language models to tackling diverse topics in computational social science, such as sentiment analysis and bias detection.

Beyond the Labels: Handling Subjectivity in Textual Analysis

Together with Laura Majer, Ana will present two published, co-authored research papers, both exploring the complex task of sentiment analysis in news headlines. Covering a topic where it is hard to stay neutral, And and Laura will focus on the broader challenges that anyone could encounter when solving a subjective problem – the key principles and pitfalls of leading data annotation projects, the connection between annotator subjectivity and model uncertainty, and the features of two contrasting machine learning paradigms in NLP: the common fine-tuning approach and the emerging in-context learning paradigm.