This project is dedicated to advancing the characterization of complex quantum many-body systems beyond current capabilities by means of pattern recognition enabled by ML techniques. Today, quantum simulators and quantum computers have entered a regime of generating large amounts of data by means of snapshot measurements. We aim to address the accompanying challenge of how to extract most information without the typically performed dimensional reduction to low-order correlation functions, which is particularly relevant for the characterization and identification of topological quantum matter such as quantum spin liquids. Concretely, we will introduce weighted wave-function networks as a novel tool enabling not only to extract patterns in an unbiased manner, but also to allow for detecting quantum entanglement of complex quantum many-body systems.