by Rebecca Schwab on November 21, 2014
Last month at Relativity Fest—our fifth annual user conference—we hosted more than 170 unique, educational sessions for our attendees. One session, Threading in Relativity: A Technical Deep Dive, covered the inner workings of the Relativity Analytics email threading engine, including all of the fielded information that’s generated, and where and how you can use that fielded data most effectively in your workflows.
If you missed the session at the show, here’s a recap of how to properly set up email threading, as well as how to tackle common obstacles.
To run email threading, you first need to create a structured analytics set. For the structured analytics set, you will need to create two saved searches, as follows:
• Parent Emails in the Document Set. This saved search should return parent emails only. Here are the suggested conditions:
Email From: is set | AND | Parent Doc ID: is not set
• Document Set to Analyze. This saved search should return parent emails and their attachments. Here are the suggested conditions:
Document is in: Parent Emails | Includes Family
Exclude documents with more than 30 MB of text
The structured analytics set for email threading will reference an analytics profile containing the settings for the email threading input fields. The fielded data will be used by the analytics engine during the email threading process. If there is no fielded email header information, you can set each field to a blank field in Relativity 8.x.
In Relativity 9, there’s a new option called “Use email header fields,” which allows you to easily run an email threading set without email header information.
When “Use email header fields” is set to Yes, all of the fields in the Email Threading Settings section of the linked analytics profile are used during email threading. When “Use email header fields” is set to No, only the following fields will be used:
• Attachment name
• Group ID
• Parent document ID
To properly thread and visualize attachments, you must have the Attachment Name, Group ID, and Parent Document ID fielded information. Additionally, it’s helpful to note that if “Use email header fields” is set to No, the resulting Email Threading Display field will be blank due to a lack of Email From and Email Subject information.
If the email header fields are being used, it’s crucial that you fill them in correctly, as inaccurate data in the email header or metadata fields can cause problems with the results.
Below are solutions to some easily addressed roadblocks you may find in your results:
1. There seem to be too many inclusive emails.
• Check the Sent Date field for a parent email in question. For a parent email, this needs to be set to the date and time. If it is set to only the date, and no time, the time will be interpreted as 12:00 a.m.—which may cause extra inclusive flags.
• Check for the Inclusive Reason of MESSAGE and compare emails using the Relativity Compare feature. It’s likely a change of text is causing the extra inclusive flag. For example, produced emails may have added Bates stamps that will cause a change of text. To handle these scenarios, write a regular expression that may be added to the structured analytics set.
2. All of the emails are incorrectly marked as DRAFT.
• A parent email will be marked as DRAFT if the Email Date Sent field is empty.
3. The attachments are not threaded with their parent emails.
• Make sure the attachment was contained in the Documents to Analyze saved search at the time the email threading set was run.
• Check the Parent ID field for the document in question. For an attachment, this needs to be set to the same value as the Document Identifier of the parent email, as shown below. If this field is empty, the attachment will not be threaded.
4. The Email Threading Display field is empty for the attachments.
• Check the Attachment Name field for the document in question. For an attachment, this needs to be set to the File Name and Extension for the given attachment (e.g., "Relativity Review Stats.xls").
5. Attached emails are being threaded separately from their parent emails.
• Attached emails need to be treated like attachments to be threaded with the parent emails. Make sure that the attached email was contained in the Documents to Analyze saved search and was not contained in the Identify Emails in Document Set saved search.
• Check the Parent Document ID field for the attached email in question. For an attached email, this field needs to be set to the same value as the Document Identifier of the parent email. If it is empty, the attached email will not be threaded as an attachment with its parent email.
For more information on the formats and requirements for the email threading input fields, see Email Threading Settings in our documentation. You can also learn more about the fields that result from email threading here. As always, don’t hesitate to contact us if you’d like more support.