On-site packaging and management

Enhancements

Only support upgrading from Luminoso Compass On-site v1.8 and above

If your Luminoso Compass version is:

Data includes everything the user creates, such as: 

There are no code incompatibilities between versions, so if you don't care about data loss, you are able to upgrade directly from v1.7 to v2.3. 

Updated product support and service levels policy

On May 11, 2020, we updated our product support and service levels policy. The policy describes Luminoso Technologies’ support levels and priorities, response times, and service levels.

Now, the updated policy also includes information specific to Luminoso on-site users. Read Luminoso’s support policies here

New hosting site for Luminoso Support Center and support ticketing

As of April 24, 2020, the new Luminoso Support Center is now live. Some notes on this update:

If you have questions about these changes, please contact your Customer Success Manager or reach out to us at support@luminoso.com.  

Upgrade to library for voting classifier training

With this release, we introduce an upgrade to the library Compass uses to train voting classifiers. The updated library is backward-compatible with existing classifiers. 

However, the log file will display warnings when a voting classifier trained with the old library is first loaded with the new code. We recommend rebuilding the classifiers to prevent unexpected or invalid classification results.

You can easily rebuild while the old version of the classifier is active. Rebuild one of two ways: 

  1. Click Rebuild Classifier in the UI. 

  2. Issue a POST request to the /api/projects/<project_id>/p/classifiers/<classifier>/rebuild/ endpoint in the API. 

As soon as the rebuilding process is complete, the old classifier is deactivated and all classification requests are handled by the new classifier.

Delete datasets from projects

Now, you can delete datasets from Compass projects. A dataset is either specified in a dataset field in your documents or in the UI using the Create Dataset button. When you create a classifier, you can specify the dataset you want to use. Once you delete that dataset, you won't be able to rebuild any classifier using it, but all classifiers remain available. 

To delete a dataset: 

  1. Log in to Compass and select the project you want to change. 

  2. Select Datasets in the Project detail sidebar. The Datasets page opens.

  3. Find the dataset you want to delete and click X in the Delete Dataset column. A confirmation window appears.

  4. Click Delete when you are ready to delete your dataset. The dataset is now deleted. Any classifier it was attached to is still available, but you can’t rebuild it using this dataset.