Authors: Brenden Grace, David W. Embley, Samuel Litster, Scott N. Woodfield, Spencer Seeger, Stephen W. Liddle
Tags: 2018, conceptual modeling
Analytical applications such as forensics, investigative journalism, and genealogy require deep data cleaning in which application-dependent semantic errors and inconsistencies are detected and resolved. To facilitate deep data cleaning, the application is modeled ontologically, and real-world crisp and fuzzy constraints are specified. Conceptual-model-based declarative specification enables rapid development and modification of the usually large number of constraints. Field tests show the prototype’s ability to detect errors and either resolve them or provide guidance for user-involved resolution. A user study also shows the value of declarative specification in deep data cleaning applications.Read the full paper here: https://link.springer.com/chapter/10.1007/978-3-030-00847-5_9
