Ágnes Horvát seeks to measure, understand, and forecast the collective behavior of networked crowds in large-scale sociotechnical systems like peer-to-peer platforms. Her current research develops empirical and theoretical methods to support creativity and predict success in culture industries, identify expressions of collective intelligence and opportunities for innovation in crowdsourcing communities, as well as detect shared misconceptions and biases in online capital markets. Her work work at the intersection of computational social science and social computing uses an interdisciplinary data-driven approach and builds on techniques from network science, machine learning, statistics, and exploratory visualization.