Review Archives

First Case for Technology Assisted Review to be Completed – eDiscovery Trends

January 21, 2013

As reported in Law Technology News by Evan Koblentz, it appears we have our first case in which predictive coding has been completed.

Last April, as reported in this blog, in Global Aerospace Inc., et al, v. Landow Aviation, L.P. dba Dulles Jet Center, et al, Virginia State Circuit Court Judge James H. Chamblin ordered that the defendants can use predictive coding for discovery in this case, despite the plaintiff’s objections that the technology is not as effective as human review. The order was issued after the defendants issued a motion requesting either that predictive coding technology be allowed in the case or that the plaintiffs pay any additional costs associated with traditional review. The defendant had an 8 terabyte data set that they were hoping to reduce to a few hundred gigabytes through advanced culling techniques.

According to the Law Technology News article, defense counsel at Schnader Harrison Segal & Lewis, and also at Baxter, Baker, Sidle, Conn & Jones, used OrcaTec’s Document Decisioning Suite technology and that OrcaTec will announce that the process is finished after plaintiff’s counsel at Jones Day did not object to the results by a recent deadline.

As reported in the article, eDiscovery analyst David Horrigan of 451 Research, expressed his surprise that Global Aerospace didn’t head in a different direction and wondered aloud why plaintiff’s counsel did not object to the results after initially objecting to the technology itself.

“It’s disappointing this issue has apparently been resolved on [plaintiff’s] missed procedural deadline,” he said. “Not unlike the predictive coding vs. keyword search debate in Kleen Products being postponed, if this court deadline has really been missed, we’ve lost an opportunity for a court ruling on predictive coding being decided on the merits.”

For more about what predictive coding is and its effectiveness, here are a couple of previous posts on the subject. For other cases where predictive coding and other technology assisted review mechanisms have been discussed, check out this year end case summary from last week.

So, what do you think? Does this pave the way for more cases to use technology assisted review? Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

2012 eDiscovery Year in Review: eDiscovery Case Law, Part 2

January 16, 2013

As we noted yesterday, eDiscoveryDaily published 98 posts related to eDiscovery case decisions and activities over the past year, covering 62 unique cases! Yesterday, we looked back at cases related to proportionality and cooperation, privilege and inadvertent disclosures, and eDiscovery cost reimbursement. Today, let’s take a look back at cases related to social media and, of course, technology assisted review(!).

We grouped those cases into common subject themes and will review them over the next few posts. Perhaps you missed some of these? Now is your chance to catch up!

SOCIAL MEDIA

Requests for social media data in litigation continue. Unlike last year, however, not all requests for social media data were granted as some requests were deemed overbroad. However, Twitter fought “tooth and nail” (unsuccessfully, as it turns out) to avoid turning over a user’s tweets in at least one case. Here are six cases related to social media data:

Class Action Plaintiffs Required to Provide Social Media Passwords and Cell Phones. Considering proportionality and accessibility concerns in EEOC v. Original Honeybaked Ham Co. of Georgia, Colorado Magistrate Judge Michael Hegarty held that where a party had showed certain of its adversaries’ social media content and text messages were relevant, the adversaries must produce usernames and passwords for their social media accounts, usernames and passwords for e-mail accounts and blogs, and cell phones used to send or receive text messages to be examined by a forensic expert as a special master in camera.

Another Social Media Discovery Request Ruled Overbroad. As was the case in Mailhoit v. Home Depot previously, Magistrate Judge Mark R. Abel ruled in Howell v. The Buckeye Ranch that the defendant’s request (to compel the plaintiff to provide her user names and passwords for each of the social media sites she uses) was overbroad.

Twitter Turns Over Tweets in People v. Harris. As reported by Reuters, Twitter has turned over Tweets and Twitter account user information for Malcolm Harris in People v. Harris, after their motion for a stay of enforcement was denied by the Appellate Division, First Department in New York and they faced a finding of contempt for not turning over the information. Twitter surrendered an “inch-high stack of paper inside a mailing envelope” to Manhattan Criminal Court Judge Matthew Sciarrino, which will remain under seal while a request for a stay by Harris is heard in a higher court.

Home Depot’s “Extremely Broad” Request for Social Media Posts Denied. In Mailhoit v. Home Depot, Magistrate Judge Suzanne Segal ruled that the three out of four of the defendant’s discovery requests failed Federal Rule 34(b)(1)(A)’s “reasonable particularity” requirement, were, therefore, not reasonably calculated to lead to the discovery of admissible evidence and were denied.

Social Media Is No Different than eMail for Discovery Purposes. In Robinson v. Jones Lang LaSalle Americas, Inc., Oregon Magistrate Judge Paul Papak found that social media is just another form of electronically stored information (ESI), stating “I see no principled reason to articulate different standards for the discoverability of communications through email, text message, or social media platforms. I therefore fashion a single order covering all these communications.”

Plaintiff Not Compelled To Turn Over Facebook Login Information. In Davids v. Novartis Pharm. Corp., the Eastern District of New York ruled against the defendant on whether the plaintiff in her claim against a pharmaceutical company could be compelled to turn over her Facebook account’s login username and password.

TECHNOLOGY ASSISTED REVIEW

eDiscovery vendors everywhere had been “waiting with bated breath” for the first case law pertaining to acceptance of technology assisted review within the courtroom. Not only did they get their case, they got a few others – and, in one case, the judge actually required both parties to use predictive coding. And, of course, there was a titanic battle over the use of predictive coding in the DaSilva Moore – easily the most discussed case of the year. Here are five cases where technology assisted review was at issue:

Louisiana Order Dictates That the Parties Cooperate on Technology Assisted Review. In the case In re Actos (Pioglitazone) Products Liability Litigation, a case management order applicable to pretrial proceedings in a multidistrict litigation consolidating eleven civil actions, the court issued comprehensive instructions for the use of technology-assisted review (“TAR”).

Judge Carter Refuses to Recuse Judge Peck in Da Silva Moore. This is only the final post of the year in eDiscovery Daily related to Da Silva Moore v. Publicis Groupe & MSL Group. There were at least nine others (linked within this final post) detailing New York Magistrate Judge Andrew J. Peck’s original opinion accepting computer assisted review, the plaintiff’s objections to the opinion, their subsequent attempts to have Judge Peck recused from the case (alleging bias) and, eventually, District Court Judge Andrew L. Carter’s orders upholding Judge Peck’s original opinion and refusing to recuse him in the case.

Both Sides Instructed to Use Predictive Coding or Show Cause Why Not. Vice Chancellor J. Travis Laster in Delaware Chancery Court – in EORHB, Inc., et al v. HOA Holdings, LLC, – has issued a “surprise” bench order requiring both sides to use predictive coding and to use the same vendor.

No Kleen Sweep for Technology Assisted Review. For much of the year, proponents of predictive coding and other technology assisted review (TAR) concepts have been pointing to three significant cases where the technology based approaches have either been approved or are seriously being considered. Da Silva Moore v. Publicis Groupe and Global Aerospace v. Landow Aviation are two of the cases, the third one is Kleen Products v. Packaging Corp. of America. However, in the Kleen case, the parties have now reached an agreement to drop the TAR-based approach, at least for the first request for production.

Is the Third Time the Charm for Technology Assisted Review? In Da Silva Moore v. Publicis Groupe & MSL Group, Magistrate Judge Andrew J. Peck issued an opinion making it the first case to accept the use of computer-assisted review of electronically stored information (“ESI”) for this case. Or, so we thought. Conversely, in Kleen Products LLC v. Packaging Corporation of America, et al., the plaintiffs have asked Magistrate Judge Nan Nolan to require the producing parties to employ a technology assisted review approach in their production of documents. Now, there’s a third case where the use of technology assisted review is actually being approved in an order by the judge.

Tune in tomorrow for more key cases of 2012 and one of the most common themes of the year!

So, what do you think? Did you miss any of these? Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

Problems with Review? It’s Not the End of the World – eDiscovery Best Practices

December 21, 2012

If you’re reading this, the Mayans were wrong… 🙂

If 2012 will be remembered for anything from an eDiscovery standpoint, it will be remembered for the arrival of Technology Assisted Review (TAR), aka Computer Assisted Review (CAR), as a court accepted method for conducting eDiscovery review. Here are a few of the recent TAR cases reported on this blog.

Many associate TAR with predictive coding, but that’s not the only form of TAR to assist with review. How the documents are organized for review can make a big difference in the efficiency of review, not only saving costs, but also improving accuracy by assigning similar documents to the same reviewer. Organizing documents with similar content into “clusters” enables each reviewer to make quicker review decisions (for example, by looking at one document to determine responsiveness and applying the same categorization to duplicates or mere variations of that first document). This also promotes consistency by enabling the same reviewer to review all similar documents in a cluster avoiding potential inadvertent disclosures where one reviewer marks a document as privileged while another reviewer fails to mark a copy of the that same document as such and that document gets produced.

Hot Neuron’s Clustify™ is an example of clustering software that examines the text in your documents, determines which documents are related to each other, and groups them into clusters, labeling each cluster with a set of keywords which provides a quick overview of the cluster, as well as a “representative document” against which all other documents in the cluster are compared.

Clustering can make review more efficient and effective for these types of documents:

Email Message Threads: The ability to group messages from a thread into a cluster enables the reviewer to quickly identify the email(s) containing the entire conversation, categorize those and either apply the same categorization to the rest or dismiss as duplicative (if so instructed).
Routine Reports: Periodic reports – such as a weekly accounts receivable report – that are generated can be grouped together in a cluster to enable a single reviewer to make a relevancy determination and quickly apply it to all documents in the cluster.
Versions of Documents: The content of each draft of a document is often similar to the previous version, so categorizing one version of the document could be quickly applied to the rest of the versions.
Published Documents: Publishing a file to Adobe PDF format generates an exact copy (from Word, Excel or other application) of the original file in content, but different in format, so these documents won’t be identified as “dupes” based on their HASH value. With clustering, those documents still get grouped together so that those non-HASH dupes are still identified and addressed.

Within the parameters of a review tool like OnDemand®, which manages the review process and delivers documents quickly and effectively for review, clustering documents can speed decision making during review, saving considerable time and review costs, yet improving consistency of document classifications.

So, what do you think? Have you used clustering software to organize documents for review? Please share any comments you might have or if you’d like to know more about a particular topic.

eDiscovery Daily will take a break for the holidays and will return on Wednesday, January 2, 2013. Happy Holidays from all of us at Cloudnine Discovery and eDiscovery Daily!

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

Baby, You Can Drive My CARRM – eDiscovery Trends

December 14, 2012

Full disclosure: this post is NOT about the Beatles’ song, but I liked the title.

There have been a number of terms applied to using technology to aid in eDiscovery review, including technology assisted review (often referred to by its acronym “TAR”) and predictive coding. Another term is Computer Assisted Review (which lends itself to the obvious acronym of “CAR”).

Now, the Electronic Discovery Reference Model (EDRM) is looking to provide an “owner’s manual” to that CAR with its new draft Computer Assisted Review Reference Model (CARRM), which depicts the flow for a successful CAR project. The CAR process depends on, among other things, a sound approach for identifying appropriate example documents for the collection, ensuring educated and knowledgeable reviewers to appropriately code those documents and testing and evaluating the results to confirm success. That’s why the “A” in CAR stands for “assisted” – regardless of how good the tool is, a flawed approach will yield flawed results.

As noted on the EDRM site, the major steps in the CARRM process are:

Set Goals

The process of deciding the outcome of the Computer Assisted Review process for a specific case. Some of the outcomes may be:

reduction and culling of not-relevant documents;
prioritization of the most substantive documents; and
quality control of the human reviewers.

Set Protocol

The process of building the human coding rules that take into account the use of CAR technology. CAR technology must be taught about the document collection by having the human reviewers submit documents to be used as examples of a particular category, e.g. Relevant documents. Creating a coding protocol that can properly incorporate the fact pattern of the case and the training requirements of the CAR system takes place at this stage. An example of a protocol determination is to decide how to treat the coding of family documents during the CAR training process.

Educate Reviewer

The process of transferring the review protocol information to the human reviewers prior to the start of the CAR Review.

Code Documents

The process of human reviewers applying subjective coding decisions to documents in an effort to adequately train the CAR system to “understand” the boundaries of a category, e.g. Relevancy.

Predict Results

The process of the CAR system applying the information “learned” from the human reviewers and classifying a selected document corpus with pre-determined labels.

Test Results

The process of human reviewers using a validation process, typically statistical sampling, in an effort to create a meaningful metric of CAR performance. The metrics can take many forms, they may include estimates in defect counts in the classified population, or use information retrieval metrics like Precision, Recall and F1.

Evaluate Results

The process of the review team deciding if the CAR system has achieved the goals of anticipated by the review team.

Achieve Goals

The process of ending the CAR workflow and moving to the next phase in the review lifecycle, e.g. Privilege Review.

The diagram does a good job of reflecting the linear steps (Set Goals, Set Protocol, Educate Reviewer and, at the end, Achieve Goals) and a circle to represent the iterative steps (Code Documents, Predict Results, Test Results and Evaluate Results) that may need to be performed more than once to achieve the desired results. It’s a very straightforward model to represent the process. Nicely done!

Nonetheless, it’s a draft version of the model and EDRM wants your feedback. You can send your comments to mail@edrm.net or post them on the EDRM site here.

So, what do you think? Does the CARRM model make computer assisted review more straightforward? Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

Percentage of eDiscovery Sanctions Cases Declining – eDiscovery Trends

December 11, 2012

According to Kroll Ontrack, the percentage of eDiscovery cases addressing sanctions “dropped by approximately ten percent” compared to 2011, while “cases addressing procedural issues more than doubled”. Let’s take a closer look at the numbers and look at some cases in each category.

As indicated in their December 4 news release, in the past year, Kroll Ontrack experts summarized 70 of the most significant state and federal judicial opinions related to the preservation, collection, review and production of electronically stored information (ESI). The breakdown of the major issues that arose in these eDiscovery cases is as follows:

Thirty-two percent (32%) of cases addressed sanctions regarding a variety of issues, such as preservation and spoliation, noncompliance with court orders and production disputes. Out of 70 cases, that would be about 22 cases addressing sanctions this past year. Here are a few of the recent sanction cases previously reported on this blog.
Twenty-nine percent (29%) of cases addressed procedural issues, such as search protocols, cooperation, production and privilege considerations. Out of 70 cases, that would be about 20 cases. Here are a few of the recent procedural issues cases previously reported on this blog.
Sixteen percent (16%) of cases addressed discoverability and admissibility issues. Out of 70 cases, that would be about 11 cases. Here are a few of the recent discoverability / admissibility cases previously reported on this blog.
Fourteen percent (14%) of cases discussed cost considerations, such as shifting or taxation of eDiscovery costs. Out of 70 cases, that would be about 10 cases. Here are a few of the recent eDiscovery costs cases previously reported on this blog.
Nine percent (9%) of cases discussed technology-assisted review (TAR) or predictive coding. Out of 70 cases, that would be about 6 cases. Here are a few of the recent TAR cases previously reported on this blog, how many did you get?

While it’s nice and appreciated that Kroll Ontrack has been summarizing the cases and compiling these statistics, I do have a couple of observations/questions about their numbers (sorry if they appear “nit-picky”):

Sometimes Cases Belong in More Than One Category: The case percentage totals add up to 100%, which would make sense except that some cases address issues in more than one category. For example, In re Actos (Pioglitazone) Products Liability Litigation addressed both cooperation and technology-assisted review, and Freeman v. Dal-Tile Corp. addressed both search protocols and discovery / admissibility. It appears that Kroll classified each case in only one group, which makes the numbers add up, but could be somewhat misleading. In theory, some cases belong in multiple categories, so the total should exceed 100%.
Did Cases Addressing Procedural Issues Really Double?: Kroll reported that “cases addressing procedural issues more than doubled”; however, here is how they broke down the category last year: 14% of cases addressed various procedural issues such as searching protocol and cooperation, 13% of cases addressed various production considerations, and 12% of cases addressed privilege considerations and waivers. That’s a total of 39% for three separate categories that now appear to be described as “procedural issues, such as search protocols, cooperation, production and privilege considerations” (29%). So, it looks to me like the percentage of cases addressing procedural issues actually dropped 10%. Actually, the two biggest category jumps appear to be discoverability and admissibility issues (2% last year to 16% this year) and TAR (0% last year to 9% this year).

So, what do you think? Has your organization been involved in any eDiscovery opinions this year? Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

When Considering Review Solutions, Don’t Forget About Self-Documentation – eDiscovery Best Practices

December 6, 2012

When evaluating eDiscovery review solutions, there are a number of features that attorneys consider as part of their selection process. For example: What searching capabilities does the solution have? How does it handle native files? How does it support annotations and redactions of images? Can it support conceptual clustering and predictive coding? But, one feature that often gets overlooked is the ability for the review solution to automatically document searching and review activities. Not only does that make it easier to identify potential issues in the process; it also facilitates the ability for attorneys to demonstrate a defensible approach to discovery to the court.

There are at least three areas where self-documentation can assist in the discovery analysis and review process:

Searching: An application, such as FirstPass®, powered by Venio FPR™, that keeps track of every search in a search history can provide assistance to attorneys to demonstrate a defensible search approach. eDiscovery searching is almost always an iterative process where you perform a search, analyze the results (often through sampling of the search results, which FirstPass also supports), then adjust the search to either add, remove or modify terms to improve recall (when responsive information is being missed) or improve precision (when the terms are overly broad and yielding way too much non-responsive information, such as the “mining” example we’ve discussed previously).

Tracking search history accomplishes two things: 1) it makes it easier to recall previous searches and “tweak” them to run a modified version of the search without starting from scratch (some searches can be really complex, so this can be a tremendous time saver) and, 2) it documents the natural iterative process of searching, facilitating the ability for attorneys to demonstrate a defensible search approach to discovery to the court, if necessary. And, if you don’t think that ever comes up, check out these case summaries here, here, here and here. Not only that, the ability to look at previous searches can be a shorten the learning curve for new users that need to conduct searches by giving them examples after which to pattern their own searches.

Tomorrow, we’ll discuss the other two areas where self-documentation can assist in the discovery analysis and review process. Let the anticipation build!

So, what do you think? How important are self-documentation features in a review solution to you? Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

The Grossman-Cormack Glossary of Technology Assisted Review – eDiscovery Resources

November 29, 2012

Do you know what a “Confidence Level” is? No, I’m not talking about Tom Brady completing football passes in coverage. How about “Harmonic Mean”? Maybe if I hum a few bars? Gaussian Calculator? Sorry, it has nothing to do with how many Tums you should eat after a big meal. No, the answer to all of these can be found in the new Grossman-Cormack Glossary of Technology Assisted Review.

Maura Grossman and Gordon Cormack are educating us yet again with regard to Technology Assisted Review (TAR) with a comprehensive glossary that defines key TAR-related terms and also provides some key case references, including EORHB, Global Aerospace, In Re: Actos:, Kleen Products and, of course, Da Silva Moore. The authors of the heavily cited article Technology-Assisted Review in E-Discovery Can Be More Effective and More Efficient Than Exhaustive Manual Review have provided a new reference document that may help many in the industry understand key TAR concepts better. Or, at least recognize key terms associated with TAR. This is version 1.01, published just this month and clearly intended to evolve over time. As the authors note in the Preamble:

“The introduction of TAR into the legal community has brought with it much confusion because different terms are being used to refer to the same thing (e.g., ‘technology assisted review,’ ‘computer-assisted review,’ ‘computer-aided review,’ ‘predictive coding,’ and ‘content based advanced analytics,’ to name but a few), and the same terms are also being used to refer to different things (e.g., ‘seed sets’ and ‘control sample’). Moreover, the introduction of complex statistical concepts, and terms-of-art from the science of information retrieval, have resulted in widespread misunderstanding and sometimes perversion of their actual meanings.

This glossary is written in an effort to bring order to chaos by introducing a common framework and set of definitions for use by the bar, the bench, and service providers. The glossary endeavors to be comprehensive, but its definitions are necessarily brief. Interested readers may look elsewhere for detailed information concerning any of these topics. The terms in the glossary are presented in alphabetical order, with all defined terms in capital letters.

In the future, we plan to create an electronic version of this glossary that will contain live links, cross references, and annotations. We also envision this glossary to be a living, breathing work that will evolve over time. Towards that end, we invite our colleagues in the industry to send us their comments on our definitions, as well as any additional terms they would like to see included in the glossary, so that we can reach a consensus on a consistent, common language relating to technology assisted review. Comments can be sent to us at mrgrossman@wlrk.com and gvcormac@uwaterloo.ca.”

Live links, with a Table of Contents, in a (hopefully soon) next iteration will definitely make this guide even more useful. Nonetheless, it’s a great resource for those of us that have bandied around these terms for some time.

So, what do you think? Will this glossary help educate the industry and help standardize use of the terms? Or will it lead to one big “Confusion Matrix”? (sorry, I couldn’t resist) Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

Louisiana Order Dictates That the Parties Cooperate on Technology Assisted Review – eDiscovery Case Law

November 19, 2012

During this Thanksgiving week, we at eDiscovery Daily thought it would be a good time to catch up on some cases we missed earlier in the year. So, we will cover a different case each day this week. Enjoy!

In the case In re Actos (Pioglitazone) Products Liability Litigation, No. 6:11-md-2299, (W.D. La. July 27, 2012), a case management order applicable to pretrial proceedings in a multidistrict litigation consolidating eleven civil actions, the court issued comprehensive instructions for the use of technology-assisted review (“TAR”).

In an order entitled “Procedures and Protocols Governing the Production of Electronically Stored Information (“ESI”) by the Parties,” U.S. District Judge Rebecca Doherty of the Western District of Louisiana set forth how the parties would treat data sources, custodians, costs, and format of production, among others. Importantly, the order contains a “Search Methodology Proof of Concept,” which governs the parties’ usage of TAR during the search and review of ESI.

The order states that the parties “agree to meet and confer regarding the use of advanced analytics” as a “document identification mechanism for the review and production of . . . data.” The parties will meet and confer to select four key custodians whose e-mail will be used to create an initial sample set, after which three experts will train the TAR system to score every document based on relevance. To quell the fears of TAR skeptics, the court provided that both parties will collaborate to train the system, and after the TAR process is completed, the documents will not only be randomly sampled for quality control, but the defendants may also manually review documents for relevance, confidentiality, and privilege.

The governance order repeatedly emphasizes that the parties are committing to collaborating throughout the TAR process and requires that they meet and confer prior to contacting the court for a resolution.

So, what do you think? Should more cases issue instructions like this? Please share any comments you might have or if you’d like to know more about a particular topic.

Case Summary Source: Applied Discovery (free subscription required). For eDiscovery news and best practices, check out the Applied Discovery Blog here.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

Searching for Email Addresses Can Have Lots of Permutations Too – eDiscovery Best Practices

November 15, 2012

Tuesday, we discussed the various permutations of names of individuals to include in your searching for a more complete result set, as well as the benefits of proximity searching (broader than a phrase search, more precise than an AND search) to search for names of individuals. Another way to identify documents associated with individuals is through their email addresses.

Variations of Email Addresses within a Domain

You may be planning to search for an individual based on their name and the email domain of their company (e.g., daustin@cloudnincloudnine.comm), but that’s not always inclusive of all possible email addresses for that individual. Email addresses for an individual’s domain might appear to be straightforward, but there might be aliases or other variations to search for to retrieve emails to and from that individual at that domain. For example, here are three of the email addresses to which I can receive email as a member of CloudNine Discovery:

daustin@cloudnincloudnine.comm;
daustin@ediscoverydaily.com (an alias to the first address to enable people to send me emails related to this blog);
support@cloudnincloudnine.comm (an Outlook contact group which includes people within our company that provide support to our applications OnDemand® and FirstPass®)

To retrieve all of the emails to and from me, you would have to include all of the above addresses (and others too). There are other variations you may need to account for, as well. Here are a couple:

Jim Smith[/O=FIRST ORGANIZATION/OU=EXCHANGE ADMINISTRATIVE GROUP (GZEJCPIG34TQEMU)/CN=RECIPIENTS/CN=JimSmith] (legacy Exchange distinguished name from old versions of Microsoft Exchange);
IMCEANOTES-Andy+20Zipper_Corp_Enron+40ECT@ENRON.com (an internal Lotus Notes representation of an email address from the Enron Data Set).

As you can see, email addresses from the business domain can be represented several different ways, so it’s important to account for that in your searching for emails for your key individuals.

Personal Email Addresses

Raise your hand if you’ve ever sent any emails from your personal email account(s) through the business domain, even if it’s to remind you of something. I suspect most of your hands are raised – I know mine is. Identifying personal email accounts for key individuals can be important for two reasons: 1) those emails within your collection may also be relevant and, 2) you may have to request additional emails from the personal email addresses in discovery if it can be demonstrated that those accounts contain relevant emails.

Searching for Email Addresses

To find all of the relevant email addresses (including the personal ones), you may need to perform searches of the email fields for variations of the person’s name. So, for example, to find emails for “Jim Smith”, you may need to find occurrences of “Jim”, “James”, “Jimmy”, “JT” and “Smith” within the “To”, “From”, “Cc” and “Bcc” fields. Then, you have to go through the list and identify the email addresses that appear to be those for Jim Smith. Any email addresses for which you’re not sure whether they belong to the individual or not (e.g., does jsmith1963@gmail.com belong to Jim Smith or Joe Smith?), you may need to retrieve and examine some of the emails to make that determination. If he uses nicknames for his personal email addresses (e.g., huggybear2012@msn.com), you should hopefully be able to identify those through emails that he sends to his business account.

In its Email Analytics module, FirstPass® makes it easy to search for email addresses for an individual – simply go to Global Email Search and type in the string to retrieve all email addresses in the collection with that string. It really streamlines the process of identifying email addresses for an individual and then reviewing those emails.

Whether or not your application simplifies that process, searching by email address is another way to identify documents pertaining to a key individual. The key is making sure your search includes all the email addresses possible for that individual.

So, what do you think? How do you handle searching for key individuals within your document collections? Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by CloudNine Discovery. eDiscoveryDaily is made available by CloudNine Discovery solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscoveryDaily should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.

Review