Abstract
One of the characteristics of big data is its internal complexity and also variety manifested in many types of datasets that are to be managed, searched, or analyzed. In their natural forms, some of the data entities are unstructured, such as texts or multimedia objects, while some are structured but too complex.
In this paper, we have investigated how visualizations of various complex datasets perform in the role of universal data representations for both human users and deep learning models. In a user study, we have evaluated several visualizations of complex relational data, where some proved their superior performance with respect to the precision and speed of classification by human users.
Moreover, the same visualizations also led to effective classification performance when used with deep learning models.