@article{504,
  keywords = {Data integrity, Measurement, Data visualization, history, Data models, Tools, Data Wrangling, data cleansing, data quality, quality metrics, data provenance, sensemaking},
  author = {Christian Bors and Theresia Gschwandtner and Silvia Miksch},
  title = {Capturing and Visualizing Provenance From Data Wrangling},
  abstract = {Data quality management and assessment play a vital role for ensuring the trust in the data and its fitness-of-use for subsequent analysis. The transformation history of a data wrangling system is often insufficient for determining the usability of a dataset, lacking information how changes affected the dataset. Capturing workflow provenance along the wrangling process and combining it with descriptive information as data provenance can enable users to comprehend how these changes affected the dataset, and if they benefited data quality. We present DQProv Explorer, a system that captures and visualizes provenance from data wrangling operations. It features three visualization components: allowing the user to explore the provenance graph of operations and the data stream, the development of quality over time for a sequence of wrangling operations applied to the dataset, and the distribution of issues across the entirety of the dataset to determine error patterns.},
  year = {2019},
  journal = {IEEE Computer Graphics and Applications},
  volume = {39},
  chapter = {61},
  number = {6},
  pages = {15},
  month = {11/2019},
  issn = {1558-1756},
  url = {https://publik.tuwien.ac.at/files/publik_282245.pdf},
  doi = {10.1109/MCG.2019.2941856},
  note = {zur Veröffentlichung angenommen},
}