5 papers
Whose Name Comes Up? III: Persona Prompting Effects in LLM-Based Scholar Recommendation
Annabella Sánchez-Guzmán, Lukas Eberhard, Denis Helic +1
Large language models (LLMs) are increasingly used as scholar recommenders, shaping who is seen as an expert in academia. Existing audits remain English-centric, single discipline,…
Reddit's Appetite: Predicting User Engagement with Nutritional Content
Gabriela Ozegovic, Thorsten Ruprechter, Denis Helic
Food communities on online platforms enjoy great popularity among social media users. Due to the far-reaching consequences of food-related content on user eating behavior, recent r…
Refactoring Detection in C++ Programs with RefactoringMiner++
Benjamin Ritz, Aleksandar Karakaš, Denis Helic
Commits often involve refactorings -- behavior-preserving code modifications aiming at software design improvements. Refactoring operations pose a challenge to code reviewers, as d…
NutriTransform: Estimating Nutritional Information From Online Food Posts
Thorsten Ruprechter, Marion Garaus, Ivo Ponocny +1
Deriving nutritional information from online food posts is challenging, particularly when users do not explicitly log the macro-nutrients of a shared meal. In this work, we present…
Large Language Models as Narrative-Driven Recommenders
Lukas Eberhard, Thorsten Ruprechter, Denis Helic
Narrative-driven recommenders aim to provide personalized suggestions for user requests expressed in free-form text such as "I want to watch a thriller with a mind-bending story, l…