VoD - Understanding Structure, Content, and Quality of a Dataset

In the age of data science analysts need to handle new data sets on a daily basis. In a first step they need to understand structure, content, and if the dataset is fit-for-use for further processing. However, getting familiar with a dataset by simply scrolling through the data in tabular form is just not feasible for these usually very large sets of data. Thus, we have designed and evaluated a Visual Analytics prototype that provides interactive visual summaries of a dataset on three different levels: the dataset level, the data attribute level, and the data value level. Our results demonstrate the usefulness of our approach and point to further research challenges.