• No results found

PRESENTED BY: Sponsored by:

N/A
N/A
Protected

Academic year: 2021

Share "PRESENTED BY: Sponsored by:"

Copied!
31
0
0

Loading.... (view fulltext now)

Full text

(1)

PRESENTED BY:

(2)

www.evolvediscovery.com

Practical Uses of Analytics in E-Discovery - A PRIMER

Jenny Le, Esq.

Vice President of Discovery Services

jle@evolvediscovery.com

E-Discovery & Ethics

Structured, Conceptual, and Predictive Analytics

(3)

www.evolvediscovery.com

Updates made to ABA Model Rules of Professional

Responsibility by ABA Commission on Ethics 20/20

recommendations and adopted by the ABA House of

Delegates

August 2012

Updates “reflect the realities of the digital age”

“technology has irrevocably changed and continues

to alter the practice of law in fundamental ways”

Technology

DOING THE RIGHT THING

(4)

www.evolvediscovery.com

Technology is also having a related impact on how

lawyers conduct investigations, engage in legal

research, advise their clients, and conduct

discovery

.

These tasks now require lawyers to have a firm grasp

on how electronic information is created, stored, and

retrieved. Lawyers need to know how to make and

respond to electronic discovery requests and to advise

their clients regarding electronic discovery obligations

… These developments highlight the importance of

keeping abreast of changes in

relevant technology

in

order to ensure that clients receive competent and

efficient legal services.

E-Discovery

Keeping abreast of the benefits and risks associated

with relevant technology is especially challenging for

e-discovery practitioners

Relevant technology is everywhere!

(5)

www.evolvediscovery.com

Lawyers representing clients in matters related to

e-discovery have a duty to remain current on technologies

relevant to litigation and technologies that enable

relevant data to be preserved, collected, processed,

reviewed, and produced accurately, efficiently and in a

manner consistent with all of the client’s legal

obligations.

Ethical Duty

Maintain a baseline understanding of technology

to leverage greater knowledge and expertise of

your experts

(6)

www.evolvediscovery.com

Reasonably

Necessary

Legal Knowledge

Skill

Thoroughness

Preparation

Reasonable

Assurance

Managerial Authority

Supervisory Authority

Monitoring

Reasonable Efforts

Prevent:

Inadvertent/

Unauthorized disclosure

or

unauthorized access

Client Information

Competence

Rule 1.1

Responsibilities Regarding Nonlawyer Assistance

Rule 5.3

Protecting Client Confidence

Rule 1.6

E-DISCOVER Y & ETHICS

(7)

www.evolvediscovery.com

Not Every Case Involves E-Discovery, but

Almost Every Case Potentially Does

Can you devise an e-discovery plan in 2 hours?

STATE ETHICS RULES

States and local bar associations

have issued

individual ethics opinions relating to attorney

competence in the areas of technology, cloud

computing, security & Ediscovery competence.

15

E-DISCOVERY & ETHICS

State Adoption

California New York Arkansas Arizona Alabama Delaware North Carolina Pennsylvania

(8)

www.evolvediscovery.com

Standing Order M10-468

Joint Electronic Discovery Submission & Order

Competence

Counsel certifies that they are sufficiently

knowledgeable in matters relating to their clients’

technological systems to discuss competently

issues relating to electronic discovery, or have

involved someone competent to address these

issues on their behalf.

E-DISCOVERY & ETHICS

(9)

www.evolvediscovery.com

E-DISCOVERY & ETHICS

California

Consult with

Expert

Mental, Emotional,

Physical Ability

Learning and Skill

Diligence

Case-by-Case Analysis

Knoweldge will depend on the Ediscovery issues and the nature of ESI involved.

E-Discovery Knowledge Options

Acquire sufficient learning and skill before performance

Associate or consult technical consultants or competent counsel Decline the client representation.

Proposed Ethics Opinion No. 11-0004

Rules and procedures, when placed in conjunction with

ever-changing technology, produce professional challenges that attorneys must meet in order to remain competent.

The duty exists in connection with e-discovery.

Competence

Depending on the factual circumstances, a lack of technological knowledge in handling e-discovery may

render an attorney ethically incompetent to handle certain litigation matters involving e-discovery.

(10)

www.evolvediscovery.com

Monetary Sanctions

No. 09-15173, 2010 WL 1418861 (Bankr. SDNY, Apr 7, 2010)

Plaintiffs GFI Acquisition LLC

was imposed monetary sanctions because

counsel was “uninformed on the detailed workings of GFI’s computer system

and email retention policies” and “ did not understand the technical depths to

which electronic discovery can sometimes go.”

Court awarded sanctions despite the fact that no spoliation or bad faith

conduct were found.

E-DISCOVERY & ETHICS

In re: A&M Florida Properties II, LLC

(11)

www.evolvediscovery.com

E-DISCOVERY & ETHICS

J-M Mfg. Co. v. McDermott Will & Emery

EDiscovery Malpractice Claim

No.BC462832 (Cal. Super. Ct., LA Cnty., June 2, 2011)

Plaintiff

J-M Manufacturing

, alleges that its former law firm breached its

professional duty of care by failing to supervise its EDiscovery vendors and

contract lawyers and, as a result, inadvertently producing thousands of

non-responsive, privileged documents to the government – not just once, but

twice.

“monitoring” is to remain aware of how nonlawyer services are

being performed

Rule 5.3

Discussion between Client and outside

counsel

Who is Monitoring?

Understand the technology being used

(12)

www.evolvediscovery.com

2009

2014

32M

189M

THE BASICS

Big Data

the last 5 years. 32 million documents in the largest case in 2009 to 189 million documents in the largest case in 2014.

Big Data

Relativity, a leading E-Discovery hosting platform estimated that the largest cases have grown 6X over

(13)

www.evolvediscovery.com

THE BASICS

E-Discovery Analytics

Structured Analytics

Email Threading, Near Duplicate Detection, Language Identification, Repeated Content Detection

Conceptual Analytics

Clustering, Categorization, Conceptual Searching, Keyword Expansion

Predictive Analytics

Predictive Coding, Assisted Review

(14)

www.evolvediscovery.com

Email Threading

Near Duplicate

Detection

Language

Identification

Repeated Content

Detection

E-DISCOVERY & ETHICS

Structured Analytics Suite

(15)

www.evolvediscovery.com

“Reduce Review”

How-To:

•  Sort in Email Thread Order

•  Email Thread suppressed review

•  Negotiate last-in-time productions protocol

•  Propagate coding up-thread, if necessary

E-DISCOVERY ANALYTICS

(16)

www.evolvediscovery.com

Understand custodian

communication:

Visually group together

emails in a way that’s easy

to understand.

Review only what you

need to:

Identify the inclusive emails in

a thread, such as the last

e m a i l i n a l e n g t h y

conversation, and avoid the

r e p e t i t i v e r e p l i e s a n d

forwards.

Keep email organized

for reviewers:

B a t c h o u t i n c l u s i v e

documents to your reviewers

and keep email threads intact.

Improve Speed & Cost

E-DISCOVERY ANALYTICS

(17)

www.evolvediscovery.com

Decision

Consistency

How-To:

•  Verify consistency of coding decisions

•  Identify Privilege (Rule 1.6)

•  Sort review batches by Near Duplicates

•  Cull documents

•  Compare opposing production to your own collection

E-DISCOVERY ANALYTICS

Near Duplicate Detection

Use Similarity to Your Advantage

Near Duplicate detection will provide information about:

the number of near-duplicate groups in your data set, the average number of documents per group,

the average percent of similarity between the documents in each group.

(18)

www.evolvediscovery.com

Understand dataset based on language prevalence

Gain Insight

Organize review based on language

Improve Workflow

173 languages detected

Flexibility

Language identification analyzes sentence structure

and punctuation. .

How does it work?

E-DISCOVERY ANALYTICS

(19)

Language

Identification

Reports

Language Identification Summaries and

Reporting can help in understanding the

(20)

Repeated Content Filter

When using analytics, long, repetitive blocks

of text—such as email footers—can cloud

your index and produce inconsistent results.

(21)

www.evolvediscovery.com

E-DISCOVERY ANALYTICS

Latent Semantic Index

that have undergone LSI will return results that are conceptually similar in meaning to the search criteria even if the results don’t share a specific word or words with the search criteria.

What is it?

Correlates semantically related terms that are latent in a collection of text – finding concepts. Concept searches, against a set of documents

LATENT

Not Obvious

Capable of Emerging

SEMANTIC

Relating to Meaning

Concepts

INDEX

Comparing

and

Organizing

Information

(22)

www.evolvediscovery.com

E-DISCOVERY & ANALYTICS

Analytics Index

Suitable Dataset

File types, sources/other?

Update

Include new data

Optimize Index

Remove non-text based documents, documents will with too much or too little text. Identify and remove repeated content.

Customize Settings

Set the minimum

occurrences, number of words, and other

customization options based on the unique needs of each data set.

(23)

www.evolvediscovery.com

Clustering groups or sorts documents based on common content. You can

cluster all documents or smaller groups.

After clusters are created batches can be created based on clusters..

Clustering

Groups issue coded docs with new documents based

on conceptual content.

Categorization

Keyword expansion allows you to see terms or keywords you might

not have originally expected.

Keyword Expansion

Applies a block of text against the database to find documents of similar

conceptual content.

This can help prioritize or help find important documents.

QC for consistency across conceptually similar documents.

Concept Search

E-DISCOVERY ANALYTICS

(24)

www.evolvediscovery.com

Improve Efficiency, Speed & Cost

Clustering

Internal Investigation

Witness Preparation

Issue Collection

Received Productions

Keyword Expansion

Internal Investigation

Concept Search

Internal Investigation

Categorization

Internal Investigation

Hot/Key Documents

Issue Collection

Reuse previous work product

E-DISCOVERY ANALYTICS

(25)

Predictive Analytics

Allowing a computer to predict properties of

documents based on samples you feed the

computer.

Review QC

Sample

Report /

Verify

Categorize

Review

Training

Sample

Complete

(26)

www.evolvediscovery.com

E-DISCOVERY ANALYTICS

Predictive Analytics Success

Harmonic Mean of Precision and Recall

How many Responsive documents did the computer predict, of all the

Responsive documents in the dataset?

Statistical sampling verfies and validates the project

How can you improve your project?

How can predictive coding enhance your litigation strategy?

Statistical Sampling

Passive or Active Learning

Strategy

Proportion of documents predicted Responsive, that are in fact, Responsive

Precision

Recall

(27)

www.evolvediscovery.com

“True Positive”

E-DISCOVERY ANALYTICS

Precision & Recall

false

negative

false

positive

true negative

true

positive

(28)

www.evolvediscovery.com

E-DISCOVERY ANALYTICS

Precision & Recall Considerations

Recall

160/195 = 82.1%

Precision

(29)

www.evolvediscovery.com

E-DISCOVERY ANALYTICS

TAR Disclosure & Cooperation

When Is the Right Time to Disclose Using TAR?

•  develop a workable search and review

methodology early rather than late in the discovery process.

•  Confine Rule 26(f) discussions to addressing

scope of discovery rather than specific processes.

How Much Disclosure about the TAR Process Is Required?

•  Parties are not required, under the FRCP, to

share predictive-coding training documents, or non-responsive sets with opposing counsel.

•  training sets and nonresponsive documents are

comparable to information contained in a document-retention policy and therefore also discoverable under the Federal Rules.

•  training sets not only go to whether a document is

responsive and discoverable but may also give

insight to how counsel prioritized documents for review and potentially also represent documents that counsel deemed central to the litigation, all of which encroach on work product.

•  ESI not responsive to the matter at hand may be

relevant in other matters in the future. The panel warned that it is not unusual for counsel to take discovery in one case in hopes of learning something about another case, a concern that could be especially relevant in unfair competition cases.

Things to consider about using TAR

•  Are you working with professionals who can

explain its rationale and process?

•  Do you understand the limitations and necessary

customizations?

•  Do you understand how much disclosure you are

really giving?

•  Do you know how to use TAR effectively?

•  How can you calculate cost savings?

Cases:

•  In re Biomet M2a Magnum Hip Implant Prods.

Liab. Litig., No. 3:12-md-2391, 2013 WL 6405156 (N.D. Ind. Aug. 21, 2013)

•  United States v. ExxonMobil Pipeline Co., No.

4:13-CV-00355, 2014 WL 2593781 (E.D. Ark. June 10, 2014)

•  Rio Tinto Plc v. Vale S.A., 2015 WL 872294

(30)

www.evolvediscovery.com

+

+

=

Email Threading

Near Duplicate Detection

Foreign Language Detection

Repeated Content Detection

Structured

Analytics

E-DISCOVERY ANALYTICS

Analytics Workflows In Concert

Clustering

Categorization

Concept Searching

Keyword Expansion

Conceptual

Analytics

Predictive Coding

Predictive

Analytics

Understand your dataset better

Streamline review

Improve speed and efficiency

(31)

Questions

References

Related documents

NOEX USA, ConocoPhillips, Total, other In 1989, NOEX USA began exploration, development, and production operations at an onshore field in Texas and offshore blocks in both deep

In parametric procedures, either associative or algorithmic, where the same medium is used for the conception, generation and representation of design ideas, different phases of

The effectiveness of the proposed Modified Artificial Bee Colony for load balancing dependent and independent tasks (MABC-LBDIID) algorithms in reducing the

Microsoft Windows Vista Wireless Network Connection Manager Broadcom IHV extensions for Windows Vista available to support Cisco Compatible Extensions. LED Activity LED Off – Radio

Simulated data is compared with the field measurements from two PV systems, and adequacy of each weather forecast is analyzed based on daily electricity production, daily

Wind power auctions in the U.K., Brazil, and South Africa, as well as coal-power auctions in India, have shown that flawed auction design and high completion risk can lead to

Another notable characteristic of debt-intolerant countries is that their debt- to-GNP ratio is much higher than that of countries with no history of default (on average, the

Once the Check Point configuration files have been con- verted to the LFA intermediate language, and the routing table has been converted into an MDL network firewall connectivity