Alexander Hoyle is an assistant professor of natural language processing at the TU Wien Faculty of Informatics, jointly appointed at the Complexity Science Hub.

His research is oriented around the development and evaluation of NLP/AI methods for computational social science. The methodological work covers both sides of the text-as-data paradigm. “Bottom-up” approaches facilitate inferences from text data, which has included the development of topic models and LLM-based techniques that help users make sense of large datasets. Conversely, “top-down” methods aim to make interpretable and verifiable measurements of complex constructs within text data. On the critical side, he has challenged the prevailing evaluation practices of widely-used methods in CSS, reorienting them toward the context of their real-world use.

Topically, he has collaborated with researchers in political science, law, economics, and mental health—although he is always open to other fields. His work has appeared in premier conferences in machine learning and natural language processing.

Prior to joining CSH, Alexander was a postdoctoral fellow at the ETH Zürich AI Center. He received his PhD in Computer Science from the University of Maryland in 2024.