This work introduces a conceptual representation for com-plex spatial arrangements of image features in large multi-media datasets. A novel data structure, termed the Spatial Event Cube (SEC), is formed from the co-occurrence ma-trices of perceptually classified features with respect to spe-cific spatial relationships. A visual thesaurus constructed using supervised and unsupervised learning techniques is used to label the image features. SECs can be used to not only visualize the dominant spatial arrangements of feature classes but also discover non-obvious configurations. SECs also provide the framework for high-level data mining tech-niques such as using the Generalized Association Rule ap-proach. Experimental results are provided for a large dataset of aerial images.