SIParCS 2018- Prasil Mainali

Prasil Mainali, Louisiana Tech University

NCAR DASH Search and Linked Data: Investigation and Implementation of Similar Assets

(Recorded Talk) | (Slides)

The various labs contributing to NCAR are producing valuable scientific datasets used in groundbreaking atmospheric science and related disciplines. To enhance discovery of and provide access to these datasets, NCAR's Data Stewardship Engineering Team (DSET) maintains and provides a searchable service called the Digital Asset Services Hub (DASH) Search. As the number of datasets being produced increases, discoverability of related datasets becomes more difficult. The goal of this project is to improve the capabilities of the DASH search to automatically find similar products of an identified NCAR dataset. We used the underlying search engine powering DASH to access NCAR products that are similar in content and name. By developing additional Python and HTML code and integrating it with existing DASH search features, we were able to develop a new "Similar Assets" section that can be displayed on the landing page of each DASH Search's asset. The "Similar Assets" section contains titles, descriptions, and types of resources for the top five NCAR assets that are related to the current dataset. These similar assets are chosen based on best match on title, tags, and scientific keywords. This implementation allows DASH Search users to discover related or similar data that are available across the NCAR organization. This new section gives scientists and users of DASH the opportunity to discover new related datasets that were previously difficult to find.

Mentors: Sophie Hou, Eric Nienhouse, Nathan Hook