O365 AND EDISCOVERY: WHAT'S NEW? - O365 and eDiscovery.pdfNEW FEATURES AND FUNCTIONS Title Status...
Transcript of O365 AND EDISCOVERY: WHAT'S NEW? - O365 and eDiscovery.pdfNEW FEATURES AND FUNCTIONS Title Status...
O365 AND EDISCOVERY: WHAT'S
NEW?
SPEAKERS
Derek Nagel, Esq.
Consultant
Epiq
(770) 390-5928
John Collins
Director, Information
Governance
Options Clearing
Corporation
312-322-6275
AGENDA
• History
• What’s new in “Standard” eDiscovery
• Advanced eDiscovery
• What’s new in Advanced eDiscovery
• Resources
EXCHANGE TO SHAREPOINT TO OFFICE 365: HISTORICAL PERSPECTIVE ON EDISCOVERY IN
OFFICE 365
Exchange 5.5
ZERO eDiscovery Capabilities
Exchange 2000
ZERO eDiscovery Capabilities
Exchange 2003
ZERO eDiscovery Capabilities
Exchange 2007
ZERO eDiscovery Capabilities
Exchange 2010
eDiscovery out of the box!
Office 365
eDiscovery out of the box
Equivio Acquisition
Processing & Analytics
Security & Compliance Center
New eDiscovery architecture
10 years: no out-of-the box eDiscovery tools 2009 to present: out-of-the-box eDiscovery tools
2009 2011 2013 2016
PowerShell
Archiving
Backup Tapes
JournalingExMerge
Clearwell, EnCase
Enterprise, etc.
Security & Compliance Center
New eDiscovery architecture
20151998
HISTORY OF O365’S EDISCOVERY TOOLS
Office 365 available 3/2011
Exchange 2010
Exchange Control Panel (ECP)
• Discovery Search
• Discovery Mailbox
• Litigation Hold
2011
Exchange 2013
Exchange Admin Center (EAC)
• Introduce In-Place Hold
• Improved indexing
• New search syntax (KQL)
February 2013
SharePoint 2013
SharePoint eDiscovery Center
• SharePoint and Exchange content
• Case model
• Filtering
February 2013
Equivio Acquisition
• January 2015 acquisition date
• Structured analytics
• Technology Assisted Review
December 2015
Security & Compliance
Center*
• Built from ground up in O365
• More scalable, faster exports
Q1 2016
July
20
15
Compliance Center
De
ce
mb
er
20
15
Protection Center
SharePoint 2010
• VERY limited eDiscovery functionality
2011
LEGACY TOOLS
Exchange Admin Center (EAC) SharePoint eDiscovery Center
SECURITY & COMPLIANCE CENTER
EDISCOVERY IN OFFICE 365
Standard eDiscovery
• Search mailboxes, OneDrive for
Business, SharePoint sites, O365
Groups, Teamso Keywords, proximity, date range, metadata
• Preserve In-Place (hold)o No disruption to custodians—they don’t know
they are on hold
o No collect or copy to preserve
o No journaling
• “Preview” preserved contento Not a review tool!
• Collect and exporto Mailbox ESI PST or MSG
o ODB and SharePoint ESI Native
Advanced eDiscovery
• Processing o OCR
o Text extraction and indexing
o MD5 hashing
o Load file creation
• Advanced Analyticso Near-duplicate detection
o Thread analysis
o Predictive coding (“relevance”)
o Themes
WHAT’S NEW?
User InterfaceNew Features
and FunctionsEnhancements
USER INTERFACE
NEW FEATURES AND FUNCTIONS
NEW FEATURES AND FUNCTIONS
Title Status Tags Added to
Roadmap
Public Disclosure
Availability Date
eDiscovery Export to compressed folder (zip file) Launched O365 09/11/2017 Q3 CY2017
eDiscovery Case Holds, Search and Export
modernization Launched Windows Desktop 09/11/2017 Q1 CY2018
eDiscovery - Export all content to archive (zip) Launched O365 01/10/2018 Q1 CY2018
eDiscovery Compliance Boundaries Launched
Information
Protection 06/27/2017 Q1 CY2018
eDiscovery Legal Hold Notices In development 09/11/2017 Q4 CY2018
Advanced eDiscovery: Search & Tagging In development O365 05/11/2018 Q4 CY2018
Microsoft Teams- eDiscovery Enhancements (calling) In development Microsoft Teams 10/24/2017 Q4 CY2017
Microsoft Teams - eDiscovery Enhancements
(Meetings) In development Microsoft Teams 10/24/2017 June CY2018
Advanced eDiscovery: Analyze non-Office 365 data Previously released
Information
Protection 11/01/2017 Q4 CY2017
Office 365 Advanced eDiscovery Optical Character
Recognition (OCR) Previously released
Information
Protection 06/08/2017 Q2 CY2017
CONTINUAL EVOLUTION
• Improved indexing
• Improved assessment
(search statistics)
• Iterative searches
• Improved reporting
• Additional content
sources subject to
tools
2011
Exchange and SharePoint Specific Tools
July 2015
Consolidated Search Against Exchange and SharePoint
December 2015 Advanced Analytics
December 2017
Import Non-O365 for Advanced eDiscovery Analysis
August 2018
Advanced eDiscovery Search and Tagging
ADVANCED EDISCOVERY IN OFFICE 365
Data Volume Increases Every Year
The majority of discovery spend is review (Rand)
EXPORT FROM 0365 REDUCED BY 60% VIA
MACHINE LEARNING IN ADVANCED EDISCOVERY
AUTOMATED TEXT ANALYSIS IN OFFICE 365
18
THEMES (CLUSTERING)
19
Identifies the conceptual
themes in a document
collection. Quickly understand
the “types” of documents.
EMAIL THREADING, NEAR DUPLICATE ANALYSIS
• Useful on almost any volume of data
• Recommended standard final step before
attorney review
20
EMAIL THREADING -DETAILS
21
PREDICTIVE CODING IN OFFICE 365
PREDICTIVE CODING
• Predictive coding
A technology-enabled process that
employs an algorithm to classify
documents
23
RELEVANCE IN ADVANCED EDISCOVERY
• Support vector machine (SVM) active learning
• System identifies text features that indicate relevant & non-relevant
documents
• Documents ranked based on presence of text features
• Built-in statistical modeling to measure effectiveness
24
THE PROCESS IS SIMPLE, BUT TAKES A FEW DAYS TO COMPLETE
25
THE EXPERT MAKES A SERIES OF YES/NO
DECISIONS
26
Simple Yes/No,
Keep it/Lose it
choice for each
document
THE FIRST SET IS A RANDOMLY SELECTED
CONTROL SET
27
Identifies Richness
at start
THE EXPERT RECEIVES FEEDBACK AFTER EACH
SET OF 40 TRAINING DOCUMENTS
• 40 document training samples
• Documents selected by the system
• Results available after each round
• Expert driven – no delay between rounds
28
PROGRESS IS UPDATED AFTER EACH ROUND OF 40 DOCUMENTS
29
PROGRESS IS UPDATED AFTER EACH ROUND OF
40 DOCUMENTS
30
TYPICALLY FEWER THAN 2,000 TRAINING
DOCUMENTS ARE REQUIRED TO COMPLETE THE
PROCESS
• System stable when F-Measure no longer improving
• System not gaining value from additional rounds of training
• Next step is to batch rank or apply the algorithm to all documents in the collection
31
THIS IS THE BEST RESULT FOR THIS TOPIC AND THESE DOCUMENTS
32
FINAL RESULTS PROVIDE GUIDANCE ON WHERE TO SET THE CUTOFF SCORE BUT THE CASE TEAM DECIDES
33
REVIEW TO RELEVANCE RATIO
34
UNDERSTANDING THE RESULTS
• RichnessThe percentage of all the documents in the population that are relevant
• RecallOf all the relevant documents in the population, the percentage that are returned by the search
• PrecisionThe percentage of documents returned by a search that are relevant
35
OPTIONS FOR VALIDATION TESTING
• Test what’s left behind
Test random selection of documents
below the cutoff score
• Test a range
Test documents in slices – ranges of
scores
• Use existing review calls
Compare against existing review calls
from another source
36
ADVANCED EDISCOVERY WORKFLOW
ADVANCED EDISCOVERY WORKFLOW
GETTING
DOCUMENTS TO
ADVANCED
EDISCOVERY
• Create and run a search associated with a case
• Prepare search results for Advanced eDiscovery
• Go to the case in Advanced eDiscovery
PROCESS &
ANALYZE DATA
IN ADVANCED
EDISCOVERY
• Must process and run ‘Analyze’ before running predictive coding
• Express Analysis – quickly processes, analyzes (threading & near-duplicate) , and exports documents
• If doing predictive coding, a two step process:– Process – extracts text and OCR images to
supply text for analytics & can import non-Office 365 data
– Analyze – Processes the text through email threading/near-duplicate analysis/Themes clustering
PREDICTIVE CODING IN RELEVANCE
• Select document set for predictive coding
(Loads)
• Configure your issue(s) (e.g.
Responsiveness)
• Assign reviewer
• Create Relevance samples from Track page
• Make review decisions from Decide page
41
EXPORT RESULTS
• Download to Azure location (for loading
into an Azure based hosting tool)
• Download to local secure location (for
loading into a service provider’s review
platform or a local on-premise hosted
review solution)
• Downloaded content is “load ready”
• Lightweight Excel available for very small
sets
42
WHAT’S NEW IN ADVANCED EDISCOVERY
IMPORT NON-OFFICE 365 CONTENT FOR ADVANCED EDISCOVERY ANALYSIS
• Introduced November 2017
• Allows for import of documents
and files to Azure storage blob
– Does NOT include PST files
• Data uploaded must be
associated with licenses O365
user with E3 and E5 or stand-
alone AeD license
SEARCH & TAGGING IN ADVANCED EDISCOVERY
• Search content within your
existing Advanced eDiscovery
case, including:
– keywords
– metadata
– Themes
– Relevance scores & tags
• Tag search results for
organization or selecting export
content
• Available Q4 CY2018 (currently
beta)
SEARCH & TAGGING IN ADVANCED EDISCOVERY
SEARCH & TAGGING IN ADVANCED EDISCOVERY
REVIEWING SEARCH RESULTS
TAG RESULTS WITH CUSTOM LABEL
EXPORT TAGGED DOCUMENTS
• Limit export to documents that
are tagged with a particular label
• Further reduce content that is
exported from Advanced
eDiscovery
SEARCH & TAGGING -LIMITATIONS
• Only the first 10,000 documents
(according to sort order) from a
given search result can be
previewed at once
• When selecting documents, you
cannot select from multiple
preview pages – you can only
select documents from the same
page, or the entire query
RESOURCES
SUPPORT.OFFICE.COM
https://support.office.com/en-us/article/Security-and-Compliance-in-Office-365-for-business-Admin-Help-7fe448f7-49bd-4d3e-919d-0a6d1cf675bb?ui=en-US&rs=en-US&ad=US
DOCS.MICROSOFT.COM
https://docs.microsoft.com/en-us/MicrosoftTeams/teams-overview
TECHNET
https://technet.microsoft.com/en-us/library/dn532171.aspx
TRUST CENTER
https://products.office.com/en-us/business/office-365-trust-center-cloud-computing-security
EXCELLENT REFERENCE
https://gumroad.com/l/O365IT
VIDEO BLOG
https://www.youtube.com/playlist?list=PLXPr7gfUMmKwn422HmCx7b7D5qh9T6frb
ROADMAP
https://products.office.com/en-us/business/office-365-roadmap
MICROSOFT IGNITE
• Annual conference
• 20k+ attendees
• Major product announcements
• Roadmaps
• eDiscovery sessions with Microsoft experts, including E.J. Bastien and Rachi Messing
– Reduce legal fees and gain insight into your data leveraging Office 365 Advanced eDiscovery
– How Microsoft Legal drives down eDiscovery costs with machine learning in Office 365
– Quickly find what’s relevant and reduce risk with intelligent eDiscovery in Office 365
To access free recorded sessions go to:
https://www.microsoft.com/en-us/ignite/default.aspx