De Cleen, ThomasBaecke, PhilippeGoedertier, Frank2025-05-082025-05-082025https://repository.vlerick.com/handle/20.500.12127/7671The 2024 Year Report for the PhD project between an energy business partner and Vlerick Business School highlights the research activities executed in 2024. Early in the year, a large dataset of customer service interactions was developed, linking conversation transcripts to customer satisfaction and recommendation scores. Using advanced Natural Language Processing (NLP) techniques, the study extracted emotional and linguistic features to analyze their impact on customer perceptions. A key focus was the application of machine learning models for Emotion Recognition in Conversations (ERC) and the integration of multimodal approaches combining textual and audio features. Mid-year, research expanded into assessing service quality through Large Language Models (LLMs) and conducting statistical analyses on the effects of emotional alignment in service interactions. The findings were consolidated into an academic manuscript, which underwent rigorous peer review and was ultimately accepted for publication in a high-impact journal. Toward the end of the year, work progressed on refining multimodal emotion recognition models and developing predictive frameworks for customer satisfaction and recommendation outcomes. Future research will explore the trade-offs between different modeling approaches and their applications in real-world service environments.enEmotion Recognition in Conversations (ERC)Customer Satisfaction ResearchNatural Language Processing (NLP)Machine Learning in BusinessPredictive Analytics in Customer Service26532315114550332