Abstract
Data exploration helps users interact with data to discover its content, explore possibilities, and find information that meets their needs. Exploration can be guided by the data’s structure, its content, or the results of calculations performed on the data—such as the distance between a value deemed interesting because it is unusual and the average of comparable values. While a valuable aid to users, data exploration requires enumerating and calculating numerous alternatives while maintaining interactivity; it is therefore a complex and demanding task. We will describe the main methods established to date; machine learning (ML) methods make a particularly interesting contribution in this area. The ability of data to “speak” to users strikes me as a key challenge for the discipline in the years to come.
