Universes
A universe holds documents segregated for a specific review effort, providing the framework to efficiently manage and track review process. Review batch creation and Predictive Coding are only possible within a universe.
The default universe is the All documents universe. It includes all documents of a matter at the time when the first user opens the matter in Axcelerate 5. If you do not create another universe, the review process is based on the All documents universe.
There are two ways of adding new documents to universes:
- Through a scope update. An update re-applies the search on which the universe is based. This may be needed if new documents are added to the matter. Whereas the All documents universe is updated automatically by default, custom universes need to be updated manually.
- Through new queries that are added to the universe scope.
Important: Working within a universe means that any associated documents that do not match the universe search criteria are ignored for the review process. They are also ignored for batch creation, even if the expansion by associated documents is configured for a batch. If associations, like family expansion, are important, this must be reflected by the search queries used for creating or enlarging the universe.
You can create as many universes as you like, and they can overlap. Numbers shown for the universe always relate to documents that are really contained in the universe. Of course, if a document is contained in two universes, the result of this document’s review can influence both universes.
All documents universe vs. custom universe with limited scope
Assume that you use the All documents universe, and it was updated, so that all documents of the matter are in the scope of this universe. When you create batches that include associated documents, or start training, all documents found by these actions are added to the respective review workflow.
Thread inclusion and training in the All documents universe
Now assume that you create a universe that contains only documents of a specific custodian. When creating the universe, you only apply the Custodian Smart Filter, but do not explicitly add any associated documents. Then, when you try to batch documents and want to include threads, only those thread members are added to the batches that are within the scope of the universe. Thread members that were not part of the universe when it was created, enlarged or updated are ignored. They will never be part of the review. The same is true for training: The system only suggests documents that are part of the universe. In the figure below, the grey background symbolizes matter documents outside the custom universe. Even if batch settings or training suggestions match these documents, they are not considered, because they are not part of the custom universe. Only documents within the white rectangle that symbolizes the custom universe are considered.
Thread inclusion and training in a custom universe that covers only a part of a matter
Use cases for custom universes
When analyzing matter documents, you may better understand an aspect of that matter if you look at a specific selection of documents. Universes help you to keep track of such a document selection. Instead of applying the same search criteria every time you want to work with the same documents, you create a universe.
These situations may be use cases for custom universes:
- If exception documents must be excluded from review, create a universe based on a search that excludes these documents.
- You want to have a closer look at the French documents in your matter. Then you can filter for the French language and use this as the search for the universe. This may be especially interesting for Predictive Coding, as training results are filtered according to the universe.
- Creating a separate universe for each custodian can deliver more precise information on the source of relevant documents — and visualize this information.