"CLARIN Tool Portal" prototype available for testing

[twagoo]

CLARIN is developing a Tool Portal, an early version of which is now available at toolportal.clarin-dev.eu

The purpose of the CLARIN Tool Portal is to make it easy for users to find and discover services and tools that are available within the CARIN infrastructure through a single, easy to use web application. The Tool Portal is not intended to replace or compete with any existing service, but rather to direct users to the ‘native’ repositories and catalogues where authoritative information can be found and access to tools or services can be obtained. Currently there is no similar service, and therefore users need to access multiple, partially overlapping web pages, catalogues, repositories etcetra with different modes of search and presentation, making this landscape more difficult to navigate than it should be for the average user of our infrastructure.

The current version of the Tool Portal incorporates metadata from a number of different locations: the repositories of various CLARIN centres, the CLARIAH-NL Tools registry, the Language Resource Switchboard and previously unpublished tool metadata gathered within CLARIAH-NL (all sources are listed on the portal’s help page). The range of metadata sources will be expanded over time to also include selected metadata subsets from the CLARIN Resource Families and the Social Sciences & Humanities Open Marketplace.

We welcome any ideas, comments and other feedback in this earlier stage. Feel free to write a comment below this post, or to send an e-mail to toolportal@clarin.eu. Keep in mind that some features are still missing, and that there are known issues. These are documented in the application’s GitHub repository.

[vincent]

How does this relate to the Resource Type tags Software, Webservices and Tools in the VLO, and the Tools category in the Resource Families?

[dieter]

Hi Vincent,

• Indeed, the goal is to include every record from the VLO that is easy to identify (e.g. Resource type = “software, web service”) - there might be more
• The tool Resource Families (listed below) are also on our wishlist to be added. But this will take a bit more time, since for the entries that are not harvested from a CLARIN repository, we will need to convert the * recently created JSON entries into CMDI.
  • Corpus Query Tools
  • Normalisation
  • Named Entity Recognition
  • Part-of-Speech Tagging and Lemmatisation
  • Tools for Sentiment Analysis

So your suggestions (and other) are very welcome. But to implement them it often costs a bit more time. Maybe also good to add that we are exploring the option to export the entries from the tool portal to the SSH Open Market Place as was recently suggested.