You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I think it'd be really helpful for people in the scientific community if there were keywords based on scientific / analytic domain (e.g., geospatial analytics, neuroscience) as opposed to the more functional computer-sciencey keywords that are there (e.g., API, wrapper, etc). Any plans to do something like this?
One idea that comes to mind is not to support this explicitly, but do improve the search algorithm in order to find packages that are conceptually related even if the search phrase doesn't explicitly show up with the package. If there's already a dependency tree that (I assume) is used in generating search results, perhaps it could be supported by an idea of project similarity based on the content of their READMEs, documentation, etc.
The text was updated successfully, but these errors were encountered:
The problem seems to be more around the UI only showing the most popular keywords in the search UI:
Those keywords should really be only the ones that are related to the projects from the search results, I wonder if we can get elasticsearch to do that automatically when faceting.
I think it'd be really helpful for people in the scientific community if there were keywords based on scientific / analytic domain (e.g., geospatial analytics, neuroscience) as opposed to the more functional computer-sciencey keywords that are there (e.g., API, wrapper, etc). Any plans to do something like this?
One idea that comes to mind is not to support this explicitly, but do improve the search algorithm in order to find packages that are conceptually related even if the search phrase doesn't explicitly show up with the package. If there's already a dependency tree that (I assume) is used in generating search results, perhaps it could be supported by an idea of project similarity based on the content of their READMEs, documentation, etc.
The text was updated successfully, but these errors were encountered: