Keyed search allows users to search on specific metadata fields with precise queries. Instead of entering general keywords into the search bar, a keyed search enables the user to target specific fields, e.g. pathogen species or health condition, to refine the relevance of results. The Discovery Portal has features that make it easier to enter and/or edit keyed searches.

You may want to use a keyed search to focus on a specific metadata field in order to deliver more targeted search results.

For example, if you want to search for datasets related to malaria, you could simply enter malaria in the search bar and this would return all datasets that have the term “malaria” anywhere in the metadata record. Alternatively, you could use a keyed search to limit the results to datasets where “malaria” matches the “health condition” metadata field. To do this, you would enter healthCondition.name:"malaria" in the search bar. Any property in the metadata schema can be used in a keyed search. This schema and list of all properties can be found on the Data Discovery Engine.

The Advanced Search tool makes it easy to create this query without needing any knowledge of the syntax or how to write out this query. Using keyed searches in this way improves the efficiency of finding relevant datasets.

As mentioned above, the Advanced Search tool makes it easy to write a custom keyed search for any metadata field. Read more about the Advanced Search tool here.

The Discovery Portal has other features that automatically perform keyed searches when users want to find datasets related to a general search query. For all datasets that have topic categories, the Discovery Portal creates buttons for each of the topic categories within the dataset record on the search page. Clicking one of these buttons returns a search for that term on the topicCategory metadata field. topicCategory is a metadata field used to classify a dataset’s primary focus or area of interest. It provides a high-level overview of the dataset’s content and makes it easier for users to search for datasets relevant to their research.

For example, if the user entered malaria in the search bar and then clicked on the “Topic Categories” button Preclinical and clinical studies in the pictured dataset, this would automatically return a search for topicCategory.name:"Preclinical and clinical studies". This feature helps users find related datasets.

Similarly, for all datasets that have keywords, the Discovery Portal creates buttons for each of the keywords within the dataset record on the dataset pages. Clicking one of these buttons returns a search for that term on the keyword metadata field. As opposed to topic categories, which are based on a controlled vocabulary and ontology, keywords are unstructured and non-standardized. However, keywords provide specific terms or characteristics of a dataset that may not be covered by the more generalized topic categories, so a user may want to use these buttons to find related datasets, as well.

For example, if the user entered malaria in the search bar, found the pictured dataset, and then clicked on the “Keywords” button Artesunate, this would automatically return a search for keywords:"artesunate".

Last updated on