It’s time to set the record straight on technology assisted review (TAR). Some people object to what they mistakenly believe is the “black box” nature of the technology, while others are hesitant to adopt an approach that they perceive as novel. This panel will dispel the myths, clarify the definitions, and shed light on the so-called “black box” of technology assisted review.
Some audience members may be surprised to learn that technology assisted review is nothing new. Search and clustering technology, for example, have been commonplace for many years. The phrase "technology assisted review" simply refers to a more efficient use of people, process, and technology that is the next evolutionary step in electronic discovery. As with other legal technologies, human expertise and a proven workflow are the keys to success. This panel will clearly explain what technology assisted review is all about and how it can be used as a tool in your practice so that you can make an informed decision about adopting it in your organization
The Codex of Business Writing Software for Real-World Solutions 2.pptx
Technology Assisted Review (TAR): Opening, Exploring and Bringing Transparency to the Black Box
1. Technology Assisted Review (TAR):
Opening, Exploring and Bringing
Transparency to the Black Box
LegalTech 2012 | February 1, 2012 | 1:45 – 3:00 PM
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York
and attribution information. January 30 – February 1, 2012
2. The content of this presentation
is the property of its authors.
Please contact Daegis
(info@daegis.com) for
acceptable use and attribution
information.
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 2
and attribution information. January 30 – February 1, 2012
3. Speakers
• David Horrigan, Esq., Analyst, eDiscovery and
Information Governance, The 451 Group
• David Leone, Esq., Director of Litigation Support
Services, Saul Ewing LLP
• Dr. Douglas W. Oard, University of Maryland
College of Information Studies
• Mike Stringer, Co-Founder & Managing
Partner, Datascope Analytics
Moderated by:
• Doug Stewart, Director of Technology, Daegis
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 3
and attribution information. January 30 – February 1, 2012
4. Many Things to Many People
• What’s in a name?
Predictive Coding
Computer / Machine / Technology Assisted Review
Auto-Classification / Tagging / Categorization
Clustering / Concept Searching / Iterative Search
• Is it Defined by the Technology Used?
• Is it Defined by a Workflow?
• Is it Defined by Human / Computer Division of Labor?
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 4
and attribution information. January 30 – February 1, 2012
5. TAR to the Rescue?
• Growing consensus that Traditional Exhaustive
Eyes-On Old-Fashioned Human Manual Linear
Document Review is no longer sufficient
Cost of Review
Increasing Data Volumes
Time Required
Risk / Quality
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 5
and attribution information. January 30 – February 1, 2012
6. TAR Acceptance / Adoption Indicators
• LTNY 2012 Superstar
9 TAR Sessions and Panels (16% of total)
• Research and Studies
• Presentations / Webinars /Podcasts
• Columns / Articles / Blogs
• Vendors and Products
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 6
and attribution information. January 30 – February 1, 2012
7. TAR Acceptance / Adoption
• “Until there is a judicial opinion approving (or even critiquing)
the use of predictive coding, counsel will just have to rely on
this article as a sign of judicial approval. In my
opinion, computer-assisted coding should be used in those
cases where it will help "secure the just, speedy, and
inexpensive" (Fed. R. Civ. P. 1) determination of cases in our e-
discovery world.”
Judge Andrew Peck, Search Forward, 10/01/2011 Law
Technology News, law.com
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 7
and attribution information. January 30 – February 1, 2012
8. TAR Acceptance / Adoption
Only thing missing seems to be widespread use.
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 8
and attribution information. January 30 – February 1, 2012
9. Perceived Obstacles to Acceptance
• Black Box Technology
• Technology can’t be explained (by attorneys)
• Is it defensible?
• Is it as good as eyes-on review?
• Lack of transparency
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 9
and attribution information. January 30 – February 1, 2012
10. What Does “Better” Mean?
D
“Better” Technique
Increasing
Success
(finding
relevant “Baseline” Technique
documents) A
C
y
B
x
Increasing Effort
(time, resources expended, etc.)
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 10
and attribution information. January 30 – February 1, 2012
11. Document Review
Case Knowledge
The
Black
Box
Unprocessed Coded
Documents Documents
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 11
and attribution information. January 30 – February 1, 2012
12. Inside Yesterday’s Black Box
Case Knowledge
Unprocessed Coded
Documents Documents
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 12
and attribution information. January 30 – February 1, 2012
13. Inside Today’s Black Box
Case Knowledge Keyword Search & Linear Review
“Reasoning”
“Representation”
“Interaction”
Unprocessed Coded
Documents Documents
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 13
and attribution information. January 30 – February 1, 2012
14. Inside Tomorrow’s Black Box
Case Knowledge Technology Assisted Review
“Reasoning”
“Representation”
“Interaction”
Unprocessed Coded
Documents Documents
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 14
and attribution information. January 30 – February 1, 2012
15. Representation
Unit
Message, document, container, etc…
“Map”
Evidence Many ways to do this
Content: what is IN it
Context: who SAID it to WHOM (and WHEN)
Description: what is SAID ABOUT it
Behavior: what is DONE WITH it
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 15
and attribution information. January 30 – February 1, 2012
16. Reasoning
Having a map improves the machine’s ability
Search to reason about documents.
Boolean queries,
example documents
Similarity in “map”
Similar documents receive similar coding
Based on content, context, behavior or description
Many ways to do this
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 16
and attribution information. January 30 – February 1, 2012
17. Interaction
User Interface
What can the user say?
What can the user see?
A bridge between
human thought and
Review Process machine reasoning
How does a user
move between
seeing and saying?
Review Workflow
How does a review team
allocate functions between
team members and the
systems that they use?
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 17
and attribution information. January 30 – February 1, 2012
18. TARgeting Your Firm
• Do Your TAR Homework
• Know Your Case
• Inform Your Legal Team
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 18
and attribution information. January 30 – February 1, 2012
19. Do Your TAR Homework
• Technology
Demo the different technologies and workflows.
Ethical Rules - Understand impact for attorneys.
• Process
Find a process you are comfortable managing.
Does it work within your larger process?
• People
Find a vendor who is knowledgeable and competent.
Educate your staff and users.
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 19
and attribution information. January 30 – February 1, 2012
20. Know Your Case
• Client Concerns
Is there a legal budget? Does this fit?
Does the client use review technologies?
• Case & Production Timelines
Training the “brain” takes time.
Murphy’s Law - Account for a new workflow.
• Document Volumes
Does the volume justify the initial setup time & expense?
Does the content & number of custodians fit?
• Review Goals
Responsiveness, Privilege, Issues
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 20
and attribution information. January 30 – February 1, 2012
21. Inform Your Legal Team
• Approach
Find a champion before you choose a case.
All cases are not equal - wait for the right opportunity.
• Position
Use analogies to current technologies.
Develop a terminology and stick to it.
• List Risks and Benefits
Explain both the upsides and the downsides.
Have a defensibility plan at the ready.
• Expectations
Define TAR’s role within the review workflow.
Be prepared to LOWER their expectations.
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 21
and attribution information. January 30 – February 1, 2012
22. Is it reasonable?
• Yes, if we followed a reasonable process.
Staffing
Training
Quality assurance
Linear Review
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 22
and attribution information. January 30 – February 1, 2012
23. Is it reasonable?
• Yes, if we followed a reasonable process.
Indexing
Query design
Sampling
•Keyword Search
Linear Review •Linear Review
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 23
and attribution information. January 30 – February 1, 2012
24. Is it reasonable?
• Yes, if we followed a reasonable process.
Rich representation
Explicit & example-based interaction
Process quality measurement
•Keyword Search Technology Assisted
Linear Review •Linear Review Review (TAR)
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 24
and attribution information. January 30 – February 1, 2012
25. Lessons Learned
• The technology is still evolving
Be flexible to emerging best practices
• Recruit an associate expert and ally
Act as a liaison and stay informed in the review
• Do not oversell the technology
Manage Expectations
• Don’t forget about legacy technologies
Leverage what you pay for in batching and review
• Any tool is only as good as the workflow
Develop one before you begin review!
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 25
and attribution information. January 30 – February 1, 2012
26. “We are stuck with technology when what we really want is
just stuff that works.”
-Douglas Adams
The Salmon of Doubt:
Hitchhiking the Galaxy One Last Time
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 26
and attribution information. January 30 – February 1, 2012
27. The content of this presentation
is the property of its authors.
Please contact Daegis
(info@daegis.com) for
acceptable use and attribution
information.
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York 27
and attribution information. January 30 – February 1, 2012
28. Thank you!
Questions?
The content of this presentation is the property of its authors.
Please contact Daegis (info@daegis.com) for acceptable use LegalTech® New York
and attribution information. January 30 – February 1, 2012
Editor's Notes
Primary: Doug Oard / David Horrigan
Primary:David Leone
Primary: Doug StewartSome indicators suggest we are far up the adoption curve 1. Trending at LTNY 2012 2. TREC Legal Track, Grossman and Cormack; Roitblat, Kershaw and Oot; 3. ESIBytes and Karl’s Occupy LTNY 4. Walk the floor– several products integrated or soon to be released
Primary: Doug StewartEven the bench and bar seem to support the use. Many advocates among advocates and judges.