A new sample for using the IBM Watson Content Hub (WCH) Delivery and Authoring Context Search API for similarity search is now available for download on GitHub.
Similarity search based on tags
This sample shows the similar (tags) search capability of IBM Watson Content Hub (WCH) contextual search service. Developers can use this technical sample to explore the WCH APIs and the data model. In addition, the sample illustrates the use of similar search provided by the contextual search service to search for images whose tags match a specified image. More details can be found in the sample’s readme.
Screenshot of sample in use
The following screenshot shows the sample in action. The “All images” list shows all the images that belong to the tenant. An area under this list displays the tags that are associated with a selected image. In addition, this action also triggers a ‘similarity’ search. This search attempts to find images that have at least one matching tag. The “query Results” list shows the results of this search query. The query sorts the results to display images with the most matching tags at the top of the list. If a user selects an image from the “Query Results” list, its tags are displayed. Matching tags are highlighted.
Resources
API Explorer reference documentation: https://developer.ibm.com/api/view/id-618
Watson Content Hub developer center: https://developer.ibm.com/wch
Watson Content Hub forum: https://developer.ibm.com/answers/smart-spaces/301/watson-content-hub.html