There are cases where there are a LOT of documents and, with them, a LOT of names to normalize. But with Chronicle, you can configure the normalization work for multiple reviewers, and then tell your team: "You get to normalize! And you get to normalize! And you get to normalize!"
From pre-loading data to breaking up the doc sets to "training" the system, you can speed up the normalization process by enabling more people to help. Here's how to get started!
Create separate document sets to normalize.
Within a normalization project, Chronicle builds the list of values to normalize by reading through the documents (and their fields) in a Relativity Saved Search.
Therefore, one way to break up the work is to create multiple Normalization projects using saved searches with distinct sets of documents. This approach will help minimize the overlap of normalized values between projects so that reviewers are focused on a smaller set of values to work with.
NOTE: Setting up the Saved Searches to achieve this depends on your data, but separating by custodians or the privilege designation tends to result in less overlap. Saved Searches obviously give you more options for filtering down on your doc list, so make it like you like it.
Train the system before bringing in more help.
Often, a core set of values needs to be normalized, and then, a longer tail of outliers will need to be looked through. By creating a project with a set of documents that cover most of those cases and having them reviewed first, Chronicle will be primed with the most common set of values to normalize and their normalized values.
At that point, you can turn any necessary additional users loose on the rest of the document set, and the only work they’ll have to do is clean up those outliers.
Pre-load normalizations and then bring in more help.
In the same way that you prime Chronicle with normalizations through training, if you’ve already dealt with these custodians in the past and have a set of normalizations you can load into Chronicle, you’ll achieve the same effect as above, wherein that past work product handles the bulk of the work. The review team only has to knock out the remaining values!
And that's how painless it is to simplify workflows with multi-reviewer normalization tasks using the powerful features available in Chronicle!
Comments