Easily the highlight of this release cycle for our dev team was having a subject matter expert in late beta work through a real-data example and say, “wait, I’m done with this dataset already?” That’s when we really knew the productivity features we’ve added to Version 1.5 can make a huge difference to our customers’ everyday handling of their data projects.
A few of the most immediately visible improvements in Curator 1.5 allow users to rapidly assign data structures to large numbers of observations and to reuse work from dataset to dataset (and even across projects.) Parsed datasets now automatically display which dimensions offer the most logical comparisons. With auto-assignment, users can then quickly create groups based on those dimensions. Curator’s new visual template editor also makes it easy to customize the webform-based Dataset Editor so users are focused on just those aspects of incoming datsets of interest to the project. This means that Curator projects can be changed and adapted on the fly, especially when new topics of interest are uncovered by the annotation team.
On the platform side of things, Curator’s data model has been redesigned to be compatible with the W3C RDF Data Cube Vocabulary, including support for the standard’s constructs of Dimensions, Groups, Values, Slices, and more. As users work, their data structures and annotations are automatically identified by URIs (Uniform Resource Identifiers) in Curator’s semantic network. Both enhancements ensure that datasets annotated in Curator can be linked to other internal or external data (present and future) as well as readily repurposed. Additional new features include tools that enhance control over Curator data export to fit the requirements of enterprise data stores and downstream analytics processes.
Curator 1.5 is a major leap forward in productivity and flexibility for Elevada’s customers. Still, we’re just as excited about features coming soon, including:
- Rapid integration and processing for very large datasets
- Enhanced import, including support for a wider variety of data formats, easier ways to take only relevant slices from large external vocabularies/ontologies, and more powerful auto-transformation (parsing out metadata and deriving annotation suggestions)
- Configurable workflow, including custom server hooks for automated execution of scripts