Ride The Lightning

A CALL FOR VOLUNTEERS FOR TREC 2009

Recently, I had the pleasure of being on an EDD conference faculty with Jason Baron, the Director of Litigation for the Office of the General Counsel at the National Archives and Records Administration. Apart from his depth of knowledge, rapid-fire speaking and rapier wit, I was struck by his enthusisam for a project which seems to be "his baby." Jason was very focused on this year's TREC 2009 and the search for volunteer participants. Here's an explanatory excerpt from the TREC website:

The Text Retrieval Conference (TREC) workshop series encourages research in information retrieval and related applications by providing a large test collection, uniform scoring procedures, and a forum for organizations interested in comparing their results. Now in its eighteenth year, the conference has become the major experimental effort in the field. Participants in the previous TREC conferences have examined a wide variety of retrieval techniques and retrieval environments, including cross-language retrieval, retrieval of web documents, multimedia retrieval, and question answering. Details about TREC can be found at the TREC web site, http://trec.nist.gov.

Make sure you read the Open Letter from Richard Braman, Ken Withers and others from the Sedona Conference located on the TREC Legal Track Page.

The Conference is conducted by the National Institute of Standards and Technology.

Jason would love to be inundated by volunteers, so please check out this wonderful opportunity! Jason is happy to answer any questions - you can reach him at jason.baron@nara.gov

E-mail: snelson@senseient.com     Phone: 703-359-0700

www.senseient.com

http://twitter.com/sharonnelsonesq

BLUE GRASS AND HOT BROWN: EDD IN KENTUCKY

Last Friday, we presented one of our most popular seminars, "Electronic Evidence in Family Law," for the Louisville Bar Association and the Kentucky chapter of the American Association of Matrimonial Lawyers. It was a sophisticated audience, almost all of whom had dealt with electronic evidence in their cases.

The topic that most interested them was spyware (not surprising, as this is generally a hot button topic everywhere). A good number of the lawyers had run into spyware in their cases. Here's one of the war stories we told:

The husband, who had remotely installed the spyware on his wife's computer, wanted to uninstall it = perhaps someone had told him that he was committing a crime. In any event, this particular brand of spyware had a mechanism for remote removal. All he had to do was enter the password he had selected in the subject line of an e-mail and send it to his estranged wife. The problem (predictably) was that he couldn't remember the password. So the wife received a comic series of e-mails, containing variants of what the the husband thought the password might be:  merryweather, merriweather, merrieweather, meriweather, and, well, you get the idea. The wife was baffled and we were greatly amused, as everyone at Sensei realized that the darn fool had forgotten the removal password.

Thank God for the stupid - they frequently keep us employed.

 

So what's a "hot brown?" It's a dish invented by the Brown Hotel, and though many restaurants serve it, we were told by our Louisville hosts that no one makes it as well. Toasted bread is topped with generous slices of turkey, covered with a mornay sauce and bacon. The Brown Hotel (a tourist attraction in and of itself), was once the place to drink and dance the night away. In the early morning, the revelers would have a traditional breakfast of ham and eggs. Wearying of the same fare, the guests asked the chef to create a new dish and the "hot brown" was born.

Our "hot brown" was marveous, but John and I both noted that this is not a dish you should eat before seeing your cardiologist!

E-mail: snelson@senseient.com     Phone: 703-359-0700

www.senseient.com

http://twitter.com/sharonnelsonesq  

GPS EVIDENCE LEADS TO MURDER CONVICTION

Shyanne Somers was just 12 years old.

She was supposed to babysit for the children of George Ford, Jr., but his wife Cindy returned home early and she wasn't needed. So George drove her home.

He told police that he had taken her, shortly after midnight, to see his horses in a nearby town. He said that she was outside of the car and that he accidentally ran over her while turning his pickup truck around. It took him two hours to get her to the hospital via a round-about route - he told authorities that he had gotten lost.

As things stood, he probably faced no worse than a reckless endangerment charge.

But then wife Cindy came forward. Suspicious that her husband was having an affair, she had placed a GPS tracking device in her husband's truck the day before Shyanne was murdered.

The GPS evidence showed that George had gone in a direction opposite from where his horses were. He spent about three hours behind an abandoned farmhouse with Shyanne. With that evidence, investigators were able to determine that Shyanne had gotten away from George and fled down the road, where George ran her over. The GPS device gave irrefutable evidence of George's route that night.

Since authorities were unable to prove that a sexual assault had taken place, it will never be known what happened to Shyanne behind the farmhouse. Only that she fled, and that George silenced the only witness by murdering her. He was convicted of second degree murder in February.

He will be sentenced in May. I hope for a very long time.

E-mail: snelson@senseient.com     Phone: 703-359-0700

http://twitter.com/sharonnelsonesq

MORE ON EDD ANALYTICS FROM EQUIVIO

Thanks to Amir Milo, the CEO of equivio, who wrote with respect to my recent posts on the use of analytics in EDD. Amir said:

"As a player in this domain (and with no wish to promote any specific product), I would like to point out that in my experience one of the most efficient “elements” in analytics is email threading. In cases where emails are grouped and ordered to threads and email sets, I frequently see savings of 50-70% in review time."

 

E-mail: snelson@senseient.com    Phone: 703-359-0700

 

http://twitter.com/sharonnelsonesq 

EDD: STARS OVER TEXAS AND LAWYERS AS PARTICLES...

Last week’s EDD conference in Houston was a great success, thanks to the hard work of the Texas State Bar staff, our speaker/host Craig Ball, and a star-studded lineup that included  Browning Marean, Ralph Losey, Jason Baron, Brandon Leatha, Judge David Waxse, Laurie Weiss, David Chaumette and Sheryl Falk.
 
We had fun throughout, but it was a particular pleasure to share the keynote stage with Browning Marean (who was uncharacteristically quiet during our Qualcomm discussion – his firm is now involved) and the redoubtable Ralph Losey, whose e-Discovery Team blog is among my most-cherished reads.

Craig and Brandon did a wonderful session on e-mail evidence that engendered a ton of audience questions – indeed, it was an inquisitive audience all day – which is any lecturer’s favorite thing.
 
A stat I didn’t know: Only 2% of U.S. law firms use overseas firms for EDD review, but of those who do, 100% describe themselves as either wholly or partially satisfied.

Another stat I didn’t know: Using Boolean searches alone, you may find (depending on many factors) anywhere between 1% and 100% of relevant evidence. However, the average is 22%.
 
Something to watch for: Next month, the Sedona Conference will release “Achieving Quality in the E-Discovery Process” – sure to be a popular download.

Judge Waxse’s solution to e-discovery disputes where one side says the other is an a****** - “Videotape the meeting, then I’ll know who is the a******* is. “ He has never, he said, had to watch a videotape. Smart judge. As he wryly noted, “Lawyers are like particles – they change when observed.”

E-mail: snelson@senseient.com        Phone: 703-359-0700

http://twitter.com/sharonnelsonesq

HOW TO GET BLOG TRAFFIC: THREE WAY SEX AND SEX WITH ALIENS

Recently, we began to get weekly blog stats from Ligit. Among other things, it shows the terms that bring folks to Ride the Lightning. As you might imagine, most of the searches are predictable:

Computer forensics, EDD, TECHSHOW, EDD vendors, forensics tools, etc. Boring, right?

But I have laughed out loud each week, because there are two search terms that are INVARIABLY on each weekly list: One is "sex with aliens" (who knew this was such a popular search term?) and the other is "three-way sex" (ok, I can see how that one is popular).

Lest you think I was writing something unprofessional, let me hasten to add that both those phrases were used in connection with real life computer cases where those things were in issue. Of course, I'm sure those who clicked on the blog were vastly disappointed to find a recounting of a computer forensics case, as I imagine they were searching for something more . . . vivid.

Now I could simply hyperlink to those cases, but what's the fun in that? You'll have to search yourselves. :-)

Obviously, I'm feeling better. Onward to Houston . . .

E-mail: snelson@senseient.com    Phone: 703-359-0700

http://twitter.com/sharonnelsonesq

HOUSTON, WE HAVE A PROBLEM

I am very happy to be lecturing at a Texas CLE in Houston on Wednesday. The problem, Houston, is a sinus infection compounded by an ear infection - and the voice of The Frog Princess. Just got started on antibiotics and hoping to be able to speak in near-human voice in a couple of days.

The keynote panel will consist of Browning Marean, Ralph Losey and myself discussing some of the interesting cases from 2008 forward, including Victor Stanley, Mancia, Qualcomm, O'Keefe, Keithley, Bray, Heriot, Gross Construction, Fannie Mae, Van Alstyne and Gutman. As we are all shy and retiring types (not), it should be a lively session.

Later, I will have the pleasure of presenting with Ralph again, as well as Judge David Waxse on privilege and professional responsibility.

John is presenting on production formats and (of course) computer forensics. Should be an excellent day of EDD CLE capped by host Craig Ball's infamously entertaining E-discovery Jeopardy. Craig missed his true calling as ringmaster in the center ring of the circus. As a master of ceremonies, he is without equal.

Now if I can just regain my voice - back to chicken soup and hot tea . . .

E-mail: snelson@senseient.com       Phone: 703-359-0700

http://twitter.com/sharonnelsonesq

EDD ANALYTICS IN PLAIN ENGLISH

As we saw yesterday, Jeffrey the lawyer was having a tough time with the technical description of EDD analytics. So Rob Robinson of Orange Legal Technologies graciously went back and tried out a new non-technical version of his earlier note, which he sent to directly to Jeffrey, who was most happy to receive it - and says he now understands.

So, for the benefit of those readers who were equally confused, here is Rob's non-technical version:

Analytics is taking the ESI available after collection and acting on it with available tools to reduce the data set as much as possible before entering into the typically more costly and time consuming phase of electronic discovery processing. Analytics should allow for the development of an understanding of the ESI prior to Meet and Confer conferences (as required) and prior to detailed planning for processing costs and review resource costs.

• Indexing is making a list of all the ESI available in the data set. 
• Filtering is getting rid of data that appears to be irrelevant to the case.
• Near Deduplication is getting rid of “like” data that appears to be available more than once in the data set.
• Sampling is looking at the ESI on a small scale to determine results that should be indicative of the entire data set.
• Search Term Scoping is seeing what search term keywords might best be recommended and used for the purpose of searching the data set for responsive files.

I am sure many other can provide even more complete and succinct definitions. Also, many people seem to use the term “Analytics” and “Early Case Assessment” interchangeably.

Additionally, while one may argue about the applicability of the term “analytics” – that is a discussion for someone else – as the term is certainly used in much of the eDiscovery literature today – hence my using the term here.

The underlying focus is that of preparing for efficient processing and review. The smaller the data set the earlier in the overall process the increasing economic (time and money) benefit “down stream”

I hope this very “informal, as we talk” explanation helps - it is an interesting discussion – and I – like many others – welcome more detailed writings on the subject – and I do not in any way shape of form pretend to be the expert – just one who is interested in doing things more efficiently in the area of eDiscovery.

Thanks Rob for investing additional time on this. Jeffrey's point, that we need to be able to make lawyers understand by simplifying our language, was well taken.

E-mail: snelson@senseient.com        Phone: 703-359-0700

http://twitter.com/sharonnelsonesq

ONCE MORE, IN ENGLISH: ANALYTICS IN ELECTRONIC DISCOVERY

In reply to yesterday's guest post by Rob Robinson, I received a comment from Jeffrey Reed, a lawyer in the Philadelphis area: http://www.linkedin.com/in/jeffreyreedesq 

What particularly struck me about his comment is that we in EDD often don't do a great job of explaining our jargon. Now, Rob was writing for colleagues, most of whom will understand what was said. But I'm sure a lot of lawyer/readers may have been confused a bit. For what it is worth, my own view is that "analytics" is part of the processing of the data - and I'm not entirely sure how the term came into use, which is why I asked Rob to explain what he means by the term.  

So in Jeffrey's own words . . .

Here's my problem. I'm a 25-year lawyer and consider myself reasonably adept at conducting conversations with my technically-trained colleagues. I've worked with all types of engineers (electrical, metallurgical, plastics, vehicle accident reconstruction, fire science, you name it). But I have a distinct inability to understand a segment of the computer guru circuit that resists using English in a manner that lends itself to being readily understood.

 

For example, Rob's article is replete with the terms "analytics" and yet he utterly fails to actually describe the "analysis" that is used. He refers to a set of five tools, namely:

 

·  Indexing

·  Filtering

·  Near De-duplication

·  Sampling

·  Search Term Scoping

 

but does not provide a scintilla of explanation how they are to be linked to perform an analysis. In fact, it almost appears that rather than an analysis, Robinson is outlining a more detailed method of processing documents. Perhaps an explanation of what he means by "sampling" and "search term scoping" might support an understanding of how analysis comes into play, but no such explanation is contained in his article. But if he is proposing an augmented version of processing, he should say so rather than promoting another piece of jargon for people to have to wade through.

 

I don't doubt that he has something valuable to say, or that a number of his other insights aren't in some way illuminating, but his main thesis gets lost in the jargon shuffle. There simply isn't any explanation or tie between what he says are the valuable "analytical tools" and the cost saving and compliance upsides. Such a tie may exist, but I would never know it by reading (and yes, re-re-re-reading) his article. Clearly, de-duping, filtering and any other sort of technique used to remove documents from the processing or review population will save money. But I think he means more than just "remove documents that don't belong in the review set" when he uses "analytics" in such a prominent manner. If he means no more than that, why use the term?

 

If you want lawyers to value what you say, say it in English.

 

***************************************

I'm going to keep that in mind Jeffrey! Thanks for writing.

 

E-mail: snelson@senseient.com      Phone: 703-359-0700

http://twitter.com/sharonnelsonesq

CONSIDERING ANALYTICS?

Just returned from ABA TECHSHOW, and now surrounded by piles and piles of documents glaring at me accusingly. So it is with great pleasure that I welcome this guest post from Rob Robinson of Orange Legal Technologies. Analytics in EDD has always seemed a tad myserious to me - not much has been written about it. I thank Rob for giving a tired blogger a break - and giving a LOT of very useful and insightful information as well.

*****************************************

Considering Analytics?   A Short, High Level, “Maybe Not Yet School Book Approved” Definition and Discussion of Electronic Discovery Analytics

While much has been written and presented on “electronic discovery analytics” within the past several years, it appears to be difficult to find both a commonly accepted concise definition of electronic discovery centric analytics as well as a contextual example of where analytics fits within the field of electronic discovery. I by no means claim that I have the “school book approved” answers for these two items, but I would share with you the following thoughts that may help you – as they have helped me - put electronic discovery centric analytics in perspective.

What is Analytics?

Although not listed in either of the EDRM or Sedona Conference glossaries, a definition that seems reasonable for describing analytics in relation to electronic discovery is as follows:

Electronic Discovery Analytics: The leveraging of Electronic Stored Information (ESI) through the use of a particular functional process to both provide context-specific insight that is actionable and to allow for the defensible reduction of ESI volume as early in the discovery process as appropriate.*

Key elements within this definition include:

• A Particular Functional Process
• The Provision of Actionable, Context Specific Insight
• A Defensible Reduction of ESI Volume

A Particular Functional Process

The definition of analytics, while written in a manner that does not narrowly define functional process (thus invalidating the definition with the introduction of new processes), alludes to at least five specific elements that form the “particular functional process” mentioned in the definition. These specific elements include but are not limited to:

• Indexing
• Filtering
• Near De-duplication
• Sampling
• Search Term Scoping

Whether employed individually or collectively, these elements appear to be some of the key elements utilized today to help provide actionable, context specific insight and to help defensibly reduce ESI volume for legal professionals.

Where does Analytics fit in the field of electronic discovery?

In seeking to understand analytics within the context of the electronic discovery process, the Waterfall Model of Electronic Discovery appears to be a new, intuitive and easily understood model that positions electronic discovery analytics at the front end of the core electronic discovery elements.

Taking into account the very well respected Electronic Discovery Reference Model, the Waterfall Model of Electronic Discovery designates analytics as an actual stage in the electronic discovery process (vs. the EDRM stage of analysis – which some view as an implied task to be conducted within each stage of the EDRM). This  analytics stage in the Waterfall Model occurs after the actual collection of ESI and prior to the actual processing of ESI and is inclusive of the “functional process” elements mentioned earlier.

 

While there are other “schools” of thought concerning analytics – this definition and positioning of analytics  as a stage within electronic discovery may help in explaining the overall benefit of analytics – benefits which consist of a reduction of time, risk and cost in the conduct of electronic discovery.

What is the benefit of Analytics?

Time: The Need for Speed

The use of analytics can help legal professionals quickly gain an understanding of potential evidence and seize the initiative in the conduct of litigation. In practical terms, the quicker a legal team can gain an understanding of available ESI, the quicker they can make early case assessments in relation to key questions to include:

• Does it appear that opposing counsel has an evidential basis for pursuing the case?
• What type of electronic discovery resources will be needed to conduct a complete document review?
• Based on FRCP 26(f)¹, what are the timeline requirements for “Meet and Confer” preparation?
• Based on potential evidence and resource requirements, will it be more cost effective to settle or pursue?

By quickly being able to answer these questions quickly, legal teams can gain the “litigation high ground” and ensure they are making informed client recommendations as early as possible in the litigation process – thus ensuring economy of effort without sacrificing the ability to achieve a desired outcome. Additionally, understanding provided by analytics can also help ensure counsel is prepared to proactively shape the direction of handling ESI during the federally mandated “Meet and Confer” process.

Risk:  More than a Board Game

Litigation is inherently rife with risk, and the complexity of discovery of ESI only increases this risk based on the intricacies of digital data, the continually growing volume of data available, and evolving ESI related law. Managing this complexity requires an understanding of what is an acceptable risk in relation to the time available and the financial resources available. In determining acceptable risk, three of the key concerns of legal professionals are:

• Will the electronic discovery approach reduce the risk of missing potentially responsive documents?
• Will the electronic discovery technologies used minimize risks associated with the transfer of data between organizations and platforms? 
• Will the electronic discovery effort be conducted in a legally defensible manner?

"The message to be taken from O'Keefe, Equity Analytics, and this opinion is that when parties decide to use a particular ESI search and retrieval methodology, they need to be aware of the literature describing the strengths and weaknesses of various methodologies." Judge Paul Grimm, District of Maryland Judge

In viewing traditional electronic discovery approaches, and with these risk considerations in mind, it appears that time available and financial resources determine the level of acceptable risk and that risk can be reduced though the use of analytics to help reduce the risk of missing potentially responsive documents.

Cost: Show Me the Money

The economics of electronic discovery are such an important factor in litigation that, in some cases, they may drive counsel recommendations as much, if not more, than the actual evidentiary position of the client. Additionally, based on the current economic conditions worldwide, many law firms and corporations have been significantly impacted financially and while litigation related to the financial crisis may be on the rise, there is also a corresponding decrease in the number of discretionary litigation efforts due to cost constraints. With this economic importance in mind, legal professionals not only want to but need to be able to conduct as thorough electronic discovery effort as possible at the lowest monetary cost possible. Key questions needing to be considered when evaluating the financial factor of electronic discovery may include:

• Based on time requirements and acceptable risk, what is the best electronic discovery approach congruent with firm and client financial resources and cost management objectives?
• Do we have the electronic discovery systems and expertise in place to conduct the electronic discovery tasks using the best electronic discovery approach congruent with client financial and cost management objectives?

Traditional electronic discovery approaches typically can cost anywhere between $40,000² to $130,000³ – exclusive of attorney review costs – to conduct the necessary electronic discovery tasks on 100GB of ESI. However, new analytics inclusive approaches can cut these costs significantly as they can perform the same tasks- albeit in a different order - for less than $30,000⁴.

Through an understanding of why the factors of time, risk, and costs factors of electronic discovery are important and how analytics impacts each of these areas, one can truly can see how the use of analytics can have an incredibly significant impact on electronic discovery.

---

¹ Federal Rules of Civil Procedure, Rule 26(f), http://www.law.cornell.edu/rules/frcp/Rule26.htm.

² @ $125/GB for Indexing/Culling, $500/GB for Processing and $67/GB/Month Hosting.

³ @ $500/GB for Processing or $1,250/GB Conversion (TIFF) and $67/GB/Month Hosting.

⁴ @ $125/GB for Indexing/Culling, $500/GB for Processing and $80/GB/Month Hosting.

* Gartner Research version of Analytics definition adjusted for electronic discovery specific activities.

Comment or thoughts on Rob's post? Please e-mail me.

E-mail: snelson@senseient.com        Phone: 703-359-07000

http://twitter.com/sharonnelsonesq