LegalTech Day 1

The annual e-discovery and legal technology show kicked off yesterday morning. WortzmanNickle were there to see what’s hot, what’s not and what’s the same.

As expected, Predictive Coding led the charge of buzz words once again this year. Many vendors offered their flavour of machine learning technology. However, some vendors appear to have realized, as we have, that Predictive Coding alone will not solve the dilemma of ever increasing e-discovery volumes and ever decreasing budgets and timelines. These forward thinking software developers are now integrating Predictive Coding into a package that includes all the tried and true e-discovery technologies, such as concept clustering, near duplication, email threading, and our trusted friend, keyword searching.

While there are many claims of unique Predictive Coding approaches, they all generally fall into one of two main categories – either quickly teach the computer up front what you’re looking for and then have it find your relevant documents, or let the computer observe as you search and find relevant documents using other methods, so that it can subtlety influence the result and present you with more likely relevant documents to review.  Both methods should theoretically end up with the same results.

Wortzman Nickle will be exploring these two approaches over the next couple of months and report in upcoming blogs and papers on the costs and benefits of each methodology.

e-Discovery Undersized

e-Discovery conjures up visions of millions of documents and cases that cost hundreds of thousands of dollars. Although these are the cases that make the headlines, the average litigation typically involves less than 10,000 records, or about one gigabyte of electronic data.

When law clerks describe “a small document” collection, the usual approach is to either print everything out for review, convert everything to tiff images, or review the native files without any special, e-discovery tools. As anyone who reads this blog (and countless others on the net) knows, these are the least efficient, and most costly methods, regardless of the collection size.

While it will not cost hundreds of thousands of dollars to deal with a small volume of electronic information, e-discovery methodologies developed for large document cases, including near duplicate and email thread grouping, statistical sampling, predictive coding, and content analysis, can be equally applied to small cases to ensure that the overall discovery cost is as low as possible.

For example, a 10,000 record collection would require a review of about 400 records in order to apply predictive coding or statistical sampling. Given the current pricing of the various predictive coding solutions, it would cost under $500 to analyse this data. Combine that with about 10 hours of lawyer time to review the 400 documents, and you would have your records all sorted and prioritized. You can then shave off the ones that are likely to be irrelevant, skip the ones you’ve already reviewed, and probably spend another 10 hours or so reviewing the rest. Compare that to the cost to print out 10 bankers boxes of paper, along with the manual (and tedious) lawyer review, and you can clearly see how this approach can save thousands of dollars and millions of brain cells.

Wortzman Nickle can add value to any size case. Call us to find out how.

The Perfect e-Discovery Solution?

The process of preserving, collecting, reviewing, and producing records is imperfect. Until all information is solely electronic and is automatically classified when created, it always will be.

Theoretically, considering that upwards of 98% of all information is electronic, it should be possible to find each and every relevant record. However, the ever-increasing volume of digital data continues to outpace our ability to efficiently and accurately deal with this information. The reality of limited time and money demands that parties compromise and accept discovery imperfection.

The problem is well documented: The amount of information subject to discovery in litigation continues to grow at almost unfathomable rates as individuals and corporations generate staggering volumes of information. In 2010, approximately 32 billion non-spam e-mails were sent every day — as compared with the 171 billion pieces of mail delivered by the U.S. Postal Service during all of 2010. In addition, social media posts, status updates, tweets, and blogs, produced from data sources such pads, pods, and clouds, all contribute to this ever increasing mass of information.

The time, burden, and costs associated with identifying and producing relevant records from mountains of information is swamping traditional discovery budgets and holding litigants in an expensive dilemma. Further complicating matters, this problem is expected to be solved in the same amount of time it took to produce documents back in the paper days.

There have been many methods developed over the years to “perfect” the e-Discovery process, such as custodian-directed collection, iterative search terms, early case-assessment, visualization, concept clustering and the newest kid on the block, predictive coding. Each of these methods has its own benefits and risks, but none produce a perfect result.

No matter how reasonable the efforts, how cooperative counsel are, or how advanced the technology is, litigants must understand that some documents will be withheld that are not privileged, some privileged documents may get produced, and some relevant documents may never see the light of day.

This is not a new problem. When paper files ruled the world, the challenge was finding critical documents that existed only within a multitude of storage boxes in some dusty warehouse. Today, the problem is almost the reverse: the chance of any single document getting lost is very small. However, having all that digital information at hand results in documents getting lost in plain sight.

Since we cannot locate, collect, and produce every relevant piece of information, what should we do? Our ethical obligations are no different than they were during the days of paper discovery. Somehow, we need to balance the requirement to produce all relevant information against the practical problems of time and expense.

There are no checklists or guidelines that lead to the perfect solution. The best way to manage these imperfections is to admit they exist, take reasonable steps to reduce them, and protect clients against them by seeking agreements that address the inevitable errors. The more transparent this process is, the more likely the parties and the courts can reach reasonable solutions. Maybe someday computers will be wise enough to save us all from ourselves, but in the mean time, the issues associated with filtering down huge amounts of information to manageable pieces will require technical know-how, foresight, cooperation and patience.

Finding what you Search for

When searching for information, it’s not the “search” that’s important, it’s what you find. According to the analyst firm Gartner, what we once knew as search, is not just search anymore. In fact, it now uses the term “information access” to include a collection of technologies to help you find information, such as:

• content classification, categorization and clustering
• fact and entity extraction
• taxonomy creation and management
• information presentation (for example visualization)

Many of the tools around extraction, classification, and categorization of records remain supplementary to the essential task of organizing information. There are three main ways in which people look for information:

• Pattern Matching – using search criteria with the same physical attributes as the sought after information, such as keyword searching. Pattern matching requires that the found information contains the words or phrases in certain parts (e.g. the title, author, content), and possibly that certain words exists close to each other (e.g. clustering).

• Semantic Web Navigation – an artificial “web” of data that allows machines to understand the semantics, or meaning, of information. Relationships between discrete pieces of information are identified, usually in some sort of visual representation.

• Classified or Categorized, that which is organized by topic browsing. – This is where we use classification taxonomies and related structured organizations of information.

While only the first approach relies exclusively on “search”, the line between search and browse (either by link or by structure) blurs more every day, as clustering and guided navigation enable new ways for lawyers to facilitate useful access to large repositories. At the end of the day, all three approaches rely heavily on metadata. Clearly, to access information properly, first you need to organize it properly.

Wortzman Nickle has spent considerable resources analysing and employing various information access technologies in an effort to maximize data analysis and review efficiencies. For more information, contact us.

e-Discovery Literacy

“The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn.” Alvin Toffler

Rapid changes in technology make this quotation particularly true in the e-discovery realm.  New approaches, different digital sources of information, new and improved technology, and the practical realities of limited resources all challenge litigators to approach each file in new and innovative ways to minimize the costs of discovery. In short, one size does not fit all.

The moral of this story is to stay flexible. Be adaptable. Don’t get tied down to one approach, one tool, or one piece of software. In the words of Toffler, “learn, unlearn and relearn” to ensure that all of the phases of discovery are conducted in the most timely and cost effective manner possible. This is simple but not easy. It requires e-discovery literacy - a commitment to stay current on new and emerging discovery approaches and technologies: early case assessment, processing, predictive coding. Your law firm and your clients will thank you.

If you don’t know what’s out there, ask your IT department, a forensic vendor, or call us. We’re always happy to discuss all things e-discovery.

Predictive Coding Demystified

Predictive coding has received a lot of attention lately as the next great magical wand in the e-discovery bag of tricks. However, as with any new technology, there are a number of different implementations and marketing claims that are confusing the whole picture of how this system can help make the e-discovery process more efficient and ultimately reduce costs.

In a nutshell, predictive coding involves the application of sophisticated artificial intelligence to permit the computer to make suggested determinations based on human interaction and the content of documents.

All predictive coding incarnations basically involve the review lawyer coding a subset of the records in the collection. The system examines the decisions made by the reviewer and identifies properties of the documents that it can use to automatically make determinations. As the reviewer continues to code documents, the system predicts what the reviewer will code. When the system’s predictions and the reviewer’s actually coding coincide (within reason), the system has learned enough to make confident predictions on its own.

Predictive coding is being applied at several stages in the e-discovery analysis and review processes:

Culling: In this mode, a lawyer who is an authority on the matter makes relevance decisions on a subset of the records. Once a sufficient number of records have been reviewed (typically a few thousand), the system applies its predictive analysis to the entire set to cull out the records most likely to be relevant. These records can then be subjected to the normal, manual review process.

Subjective Coding: The predictive coding system examines the subjective coding decisions made by lawyers as they manually review records. When a sufficient number of records have been reviewed, the system will start to make coding suggestions for subsequent records to assist the lawyers.

Review Quality Control: Along the same lines as predictive subjective coding, the system uses the subjective coding decisions made by lawyers to predict how documents should be coded. However, instead of suggesting codes for un-reviewed records, the system will apply the predictions to all manually coded records and identify those records where its predictions and the actually coding diverge. This will enable reviewers to zero in on documents that may not be coded correctly.

Prioritization of Records for Review: Predictive coding can also be used to prioritize records in a review. Once a sufficient number of records have been manually reviewed and coded, the system can group un-reviewed documents based on its coding predictions. The review project manager can then group all documents likely to be coded relevant, for instance, and assign these to be reviewed first.

Predictive coding technology is also being considered in several electronic  records management solutions to permit automatic classification of records, removing the burden from individual users.

This technology is being incorporated into more and more e-Discovery software systems, and may soon become a standard way to cull and review electronic data.

For more information on this technology and other cutting-edge e-discovery solutions, contact us.

Sophisticated software can help lawyers, not replace them

On March 5^th, the New York Times published an article entitled “Armies of Expensive Lawyers, Replaced by Cheaper Software” which discussed the “new e-discovery software that can analyze millions of documents in a fraction of the time, and at a fraction of the cost consumed by human lawyers, even deducing patterns of behaviour”.   It discussed the explosion of electronically stored information, the technology used to analyze that data and how all this has disrupted the legal job market.

The article provided a clear and concise overview of the new systems available to assist legal teams.  However, the conclusion that these technologies will replace “expensive lawyers” misses the mark. All of the technologies mentioned in the article require a combination of machine and human interaction in order to operate. Humans have to “teach” the computer to identify relevant information. As with any educational process, the more highly skilled the teacher, the better the lesson will be. What the technologies will replace are lawyers working at very basic levels without a strong understanding of the case.

Although the new technologies will force lawyers to learn new ways to approach document discovery, the technologies are just one cog in the legal machine. Lawyers still need to understand the content of the documents to build their case. The new systems will help lawyers to zero in on the documents containing the relevant content.

No matter how sophisticated a computer system gets, it can’t make subjective evaluations. The legal process is not black and white – it’s ultimately based on judgement and inference. As a result, it will always require highly skilled talent.

Live From Applied Discovery’s “Understanding Proportionality”

Tuesday morning, Lexis Nexis sponsored a panel discussion on proportionality, discovery plans, and the effects that the new Ontario rules are having on the way litigation is being carried out. The panel consisted of Master Calum Macleod, Kelly Freidman of Ogilvy Renault, and our very own Susan Nickle.

Proportionality was described as one component of a set of rules designed to encourage a cultural change in the legal community. It is no longer acceptable to proceed unilaterally – parties must come together sooner and communicate more often, in order establish a real dialog and focus on the issues at the beginning of the matter, not at the end.

The new rules, and particularly the requirement for a discovery plan, are leading lawyers to develop a better understanding of technology. To be sure, most lawyers will not become techno-geeks, but it is important for lawyers to understand how electronic information is stored and where it likely resides. It is equally important for lawyers to appreciate that discovery has not changed just because documents are stored electronically – in the end, the case will hinge on same handful of documents. The only difference is that those documents will fit on a CD rather than in a banker’s box.

There were many questions from the approximately 90 people in attendance, including dialogue about  the concept of proportionality forcing litigants to think of alternative forms of proof, as some traditional forms of proof may be too cost-prohibitive.

The seminar clearly illustrated that the new rules are having an influence, and will continue to shape litigation into a more manageable, cost effective tool to resolve disputes.

Nickle reporting – Live from LegalTech 2010

Nickle from New York with Andre, Wortzman Nickle’s Senior E-discovery Analyst and Project Manager at North America’s largest legal technology conference and trade show.  

This overwhelming spectacle (this year, complete with demonstrators protesting GB fees in front of the venue!) offers lawyers, forensic vendors, records managers and other e-discovery providers the opportunity to network, attend education sessions, and scope out “the next big thing” in the technology exhibit areas.  Breakfasts, lunches and dinners with our colleagues and friends is our favorite way to keep abreast of new technology.

While we are trying to look at everything, our focus this year rests at both ends of the e-discovery spectrum – Early Case Assessment and Review tools.  We are always searching for faster, cheaper and more efficient solutions. There is amazing software available – stay tuned for our conclusions!

More from Wortzman on Early Case Assessment

Susan Nickle and I have spent a busy week meeting with vendors and assessing litigation support tools on behalf of several clients.  The new built-in features to several of these tools allow organizations and law firms to conduct their own early case assessment in-house. This became the focus of many of the meetings we had this week.  This really ties into Nickle’s post last week with respect to in-sourcing and how much of the e-discovery process should be conducted in-house by large Canadian organizations.  As the tools are developing so rapidly, we see many ways for our clients to put themselves in a position to conduct early case assessment efficiently and in a very cost effective manner.

 Despite an initial collection of hundreds of thousands of e-mails for review, early case assessment tools have allowed us to manipulate our searches and the data to cull the collections down to very manageable review sizes.  Coupling that with the review tools that allow for clustering, threading, boolean and other types of searches, we are identifying manageable review sets of data that can be triaged in a matter of days.  Trial counsels are then able conduct a serious assessment of their case.

 Susan and I  continue the quest to find the best tools to allow our clients to manage their e-discovery reviews, both in conjunction with external support and in-house.  The reaction from our clients has been overwhelmingly positive as they see the results of the early case assessment work