Text analysis

Research focus

Automatic text analysis is a systematic approach of examining written content to uncover hidden patterns, feelings and themes. This method can be applied to a variety of electronic documents, from literature to corporate reports. Our lab focuses, among others, on social media discourse, as social media platforms with their dynamic interactions offer an insight into the ever-evolving individual and collective consciousness of our time. Furthermore, our research investigates the objectivity of scientific communication and examines whether emotional, attention-grabbing language has increasingly found its way into scientific articles in recent decades.

Our research tools are based on three main methods. First, sentiment analysis enables us to recognize the emotional nuances in the text. Then, through topic modeling and powerful text embedding, we categorize and understand the predominant themes in large amounts of content. Finally, dictionary-based tools, as well as Large Language Models (LLMs), allow us to dive deeper and quantify specific linguistic features that may otherwise go unnoticed.