Time-series data are prevalent across various scientific disciplines. CompEngine, a web platform, is introduced as a self-organizing library designed to facilitate interdisciplinary connections between time series. It uses a canonical feature-based representation to place time series in a common feature space, enabling users to upload data and explore similar datasets, receiving alerts for future matches. Unlike metadata-organized databases, CompEngine promotes data sharing by connecting researchers based on data structure, fostering collaboration between experimental and theoretical scientists. Its extensive library also allows for comprehensive time-series algorithm characterization across diverse data types.

Ben D. Fulcher, Carl H. Lubba, Sarab S. Sethi, Nick S. Jones