e-discovery 2.0

2012 is the Year of the Dragon – which is fitting, since no other Chinese Zodiac sign represents the promise, challenge, and evolution of predictive coding technology more than the Dragon.  The few who have embraced predictive coding technology exemplify symbolic traits of the Dragon that include being unafraid of challenges and willing to take risks.  In the legal profession, taking risks typically isn’t in a lawyer’s DNA, which might explain why predictive coding technology has seen lackluster adoption among lawyers despite the hype.  This blog explores the promise of predictive coding technology, why predictive coding has not been widely adopted in eDiscovery, and explains why 2012 is likely to be remembered as the year of predictive coding.

What is predictive coding?

Predictive coding refers to machine learning technology that can be used to automatically predict how documents should be classified based on limited human input.  In litigation, predictive coding technology can be used to rank and then “code” or “tag” electronic documents based on criteria such as “relevance” and “privilege” so organizations can reduce the amount of time and money spent on traditional page by page attorney document review during discovery.

Generally, the technology works by prioritizing the most important documents for review by ranking them.  In addition to helping attorneys find important documents faster, this prioritization and ranking of documents can even eliminate the need to review documents with the lowest rankings in certain situations. Additionally, since computers don’t get tired or day dream, many believe computers can even predict document relevance better than their human counterparts.

Why hasn’t predictive coding gone mainstream yet?

Given the promise of faster and less expensive document review, combined with higher accuracy rates, many are perplexed as to why predictive coding technology hasn’t been widely adopted in eDiscovery.  The answer really boils down to one simple concept – a lack of transparency.

Difficult to Use

First, early predictive coding tools attempt to apply a complicated new technological approach to a document review process that has traditionally been very simple.  Instead of relying on attorneys to read each and every document to determine relevance, the success of today’s predictive coding technology typically depends on review decisions input into a computer by one or more experienced senior attorneys.  The process commonly involves a complex series of steps that include sampling, testing, reviewing, and measuring results in order to fine tune an algorithm that will eventually be used to predict the relevancy of the remaining documents.

The problem with early predictive coding technologies is that the majority of these complex steps are done in a ‘black box’.  In other words, the methodology and results are not always clear, which increases the risk of human error and makes the integrity of the electronic discovery process difficult to defend.  For example, the methodology for selecting a statistically relevant sample is not always intuitive to the end user.  This fundamental problem could result in improper sampling techniques that could taint the accuracy of the entire process.  Similarly, the process must often be repeated several times in order to improve accuracy rates.  Even if accuracy is improved, it may be difficult or impossible to explain how accuracy thresholds were determined or to explain why coding decisions were applied to some documents and not others.

Accuracy Concerns

Early predictive coding tools also tend to lack transparency in the way the technology evaluates the language contained in each document.  Instead of evaluating both the text and metadata fields within a document, some technologies actually ignore document metadata.  This omission means a privileged email sent by a client to her attorney, Larry Lawyer, might be overlooked by the computer if the name “Larry Lawyer” is only part of the “recipient” metadata field of the document and isn’t part of the document text.  The obvious risk is that this situation could lead to privilege waiver if it is inadvertently produced to the opposing party.

Another practical concern is that some technologies do not allow reviewers to make a distinction between relevant and non-relevant language contained within individual documents.  For example, early predictive coding technologies are not intelligent enough to know that only the second paragraph on page 95 of a 100-page document contains relevant language.  The inability to discern what language  led to the determination that the document is relevant could skew results when the computer tries to identify other documents with the same characteristics.  This lack of precision increases the likelihood that the computer will retrieve an over-inclusive number of irrelevant documents.  This problem is generally referred to as ‘excessive recall,’ and it is important because this lack of precision increases the number of documents requiring manual review which directly impacts eDiscovery cost.

Waiver & Defensibility

Perhaps the biggest concern with early predictive coding technology is the risk of waiver and concerns about defensibility.  Notably, there have been no known judicial decisions that specifically address the defensibility of these new technology tools even though some in the judiciary, including U.S. Magistrate Judge Andrew Peck, have opined that this kind of technology should be used in certain cases.

The problem is that today’s predictive coding tools are difficult to use, complicated for the average attorney, and the way they work simply isn’t transparent.  All these limitations increase the risk of human error.  Introducing human error increases the risk of overlooking important documents or unwittingly producing privileged documents.  Similarly, it is difficult to defend a technological process that isn’t always clear in an era where many lawyers are still uncomfortable with keyword searches.  In short, using black box technology that is difficult to use and understand is perceived as risky, and many attorneys have taken a wait-and-see approach because they are unwilling to be the guinea pig.

Why is 2012 likely to be the year of predictive coding?

The word transparency may seem like a vague term, but it is the critical element missing from today’s predictive coding technology offerings.  2012 is likely to be the year of predictive coding because improvements in transparency will shine a light into the black box of predictive coding technology that hasn’t existed until now.  In simple terms, increasing transparency will simplify the user experience and improve accuracy which will reduce longstanding concerns about defensibility and privilege waiver.

Ease of Use

First, transparent predictive coding technology will help minimize the risk of human error by incorporating an intuitive user interface into a complicated solution.  New interfaces will include easy-to-use workflow management consoles to guide the reviewer through a step-by-step process for selecting, reviewing, and testing data samples in a way that minimizes guesswork and confusion.  By automating the sampling and testing process, the risk of human error can be minimized which decreases the risk of waiver or discovery sanctions that could result if documents are improperly coded.  Similarly, automated reporting capabilities make it easier for producing parties to evaluate and understand how key decisions were made throughout the process, thereby making it easier for them to defend the reasonableness of their approach.

Intuitive reports also help the producing party measure and evaluate confidence levels throughout the testing process until appropriate confidence levels are achieved.  Since confidence levels can actually be measured as a percentage, attorneys and judges are in a position to negotiate and debate the desired level of confidence for a production set rather than relying exclusively on the representations or decisions of a single party.  This added transparency allows the type of cooperation between parties called for in the Sedona Cooperation Proclamation and gives judges an objective tool for evaluating each party’s behavior.

Accuracy & Efficiency

2012 is also likely to be the year of transparent predictive coding technology because technical limitations that have impacted the accuracy and efficiency of earlier tools will be addressed.  For example, new technology will analyze both document text and metadata to avoid the risk that responsive or privileged documents are overlooked.  Similarly, smart tagging features will enable reviewers to highlight specific language in documents to determine a document’s relevance or non-relevance so that coding predictions will be more accurate and fewer non-relevant documents will be recalled for review.

Conclusion - Transparency Provides Defensibility

The bottom line is that predictive coding technology has not enjoyed widespread adoption in the eDiscovery process due to concerns about simplicity and accuracy that breed larger concerns about defensibility.  Defending the use of black box technology that is difficult to use and understand is a risk that many attorneys simply are not willing to take, and these concerns have deterred widespread adoption of early predictive coding technology tools.  In 2012, next generation transparent predictive coding technology will usher in a new era of computer-assisted document review that is easy to use, more accurate, and easier to defend. Given these exciting technological advancements, I predict that 2012 will not only be the year of the dragon, it will also be the year of predictive coding.

The New Year has now dawned and with it, the certainty that 2012 will bring new developments to the world of eDiscovery.  Last month, we spotlighted some eDiscovery trends for 2012 that we feel certain will occur in the near term.  To understand how these trends will play out, it is instructive to review some of the top eDiscovery cases from 2011.  These decisions provide a roadmap of best practices that the courts promulgated last year.  They also spotlight the expectations that courts will likely have for organizations in 2012 and beyond.

Issuing a Timely and Comprehensive Litigation Hold

Case: E.I. du Pont de Nemours v. Kolon Industries (E.D. Va. July 21, 2011)

Summary: The court issued a stiff rebuke against defendant Kolon Industries for failing to issue a timely and proper litigation hold.  That rebuke came in the form of an instruction to the jury that Kolon executives and employees destroyed key evidence after the company’s preservation duty was triggered.  The jury responded by returning a stunning $919 million verdict for DuPont.

The spoliation at issue occurred when several Kolon executives and employees deleted thousands emails and other records relevant to DuPont’s trade secret claims.  The court laid the blame for this destruction on the company’s attorneys and executives, reasoning they could have prevented the spoliation through an effective litigation hold process.  At issue were three hold notices circulated to the key players and data sources.  The notices were all deficient in some manner.  They were either too limited in their distribution, ineffective since they were prepared in English for Korean-speaking employees, or too late to prevent or otherwise ameliorate the spoliation.

The Lessons for 2012: The DuPont case underscores the importance of issuing a timely and comprehensive litigation hold notice.  As DuPont teaches, organizations should identify what key players and data sources may have relevant information.  A comprehensive notice should then be prepared to communicate the precise hold instructions in an intelligible fashion.  Finally, the hold should be circulated immediately to prevent data loss.

Organizations should also consider deploying the latest technologies to help effectuate this process.  This includes an eDiscovery platform that enables automated legal hold acknowledgements.  Such technology will allow custodians to be promptly and properly apprised of litigation and thereby retain information that might otherwise have been discarded.

Another Must-Read Case: Haraburda v. Arcelor Mittal U.S.A., Inc. (D. Ind. June 28, 2011)

Suspending Document Retention Policies

Case: Viramontes v. U.S. Bancorp (N.D. Ill. Jan. 27, 2011)

Summary: The defendant bank defeated a sanctions motion because it modified aspects of its email retention policy once it was aware litigation was reasonably foreseeable.  The bank implemented a retention policy that kept emails for 90 days, after which the emails were overwritten and destroyed.  The bank also promulgated a course of action whereby the retention policy would be promptly suspended on the occurrence of litigation or other triggering event.  This way, the bank could establish the reasonableness of its policy in litigation.  Because the bank followed that procedure in good faith, it was protected from court sanctions under the Federal Rules of Civil Procedure 37(e) “safe harbor.”

The Lesson for 2012: As Viramontes shows, an organization can be prepared for eDiscovery disputes by timely suspending aspects of its document retention policies.  By modifying retention policies when so required, an organization can develop a defensible retention procedure and be protected from court sanctions under Rule 37(e).

Coupling those procedures with archiving software will only enhance an organization’s eDiscovery preparations.  Effective archiving software will have a litigation hold mechanism, which enables an organization to suspend automated retention rules.  This will better ensure that data subject to a preservation duty is actually retained.

Another Must-Read Case: Micron Technology, Inc. v. Rambus Inc., 645 F.3d 1311 (Fed. Cir. 2011)

Managing the Document Collection Process

Case: Northington v. H & M International (N.D.Ill. Jan. 12, 2011)

Summary: The court issued an adverse inference jury instruction against a company that destroyed relevant emails and other data.  The spoliation occurred in large part because legal and IT were not involved in the collection process.  For example, counsel was not actively engaged in the critical steps of preservation, identification or collection of electronically stored information (ESI).  Nor was IT brought into the picture until 15 months after the preservation duty was triggered. By that time, rank and file employees – some of whom were accused by the plaintiff of harassment – stepped into this vacuum and conducted the collection process without meaningful oversight.  Predictably, key documents were never found and the court had little choice but to promise to inform the jury that the company destroyed evidence.

The Lesson for 2012: An organization does not have to suffer the same fate as the company in the Northington case.  It can take charge of its data during litigation through cooperative governance between legal and IT.  After issuing a timely and effective litigation hold, legal should typically involve IT in the collection process.  Legal should rely on IT to help identify all data sources – servers, systems and custodians – that likely contain relevant information.  IT will also be instrumental in preserving and collecting that data for subsequent review and analysis by legal.  By working together in a top-down fashion, organizations can better ensure that their eDiscovery process is defensible and not fatally flawed.

Another Must-Read Case: Green v. Blitz U.S.A., Inc. (E.D. Tex. Mar. 1, 2011)

Using Proportionality to Dictate the Scope of Permissible Discovery

Case: DCG Systems v. Checkpoint Technologies (N.D. Ca. Nov. 2, 2011)

The court adopted the new Model Order on E-Discovery in Patent Cases recently promulgated by the U.S. Court of Appeals for the Federal Circuit.  The model order incorporates principles of proportionality to reduce the production of email in patent litigation.  In adopting the order, the court explained that email productions should be scaled back since email is infrequently introduced as evidence at trial.  As a result, email production requests will be restricted to five search terms and may only span a defined set of five custodians.  Furthermore, email discovery in DCG Systems will wait until after the parties complete discovery on the “core documentation” concerning the patent, the accused product and prior art.

The Lesson for 2012: Courts seem to be slowly moving toward a system that incorporates proportionality as the touchstone for eDiscovery.  This is occurring beyond the field of patent litigation, as evidenced by other recent cases.  Even the State of Utah has gotten in on the act, revising its version of Rule 26 to require that all discovery meet the standards of proportionality.  While there are undoubtedly deviations from this trend (e.g., Pippins v. KPMG (S.D.N.Y. Oct. 7, 2011)), the clear lesson is that discovery should comply with the cost cutting mandate of Federal Rule 1.

Another Must-Read Case: Omni Laboratories Inc. v. Eden Energy Ltd [2011] EWHC 2169 (TCC) (29 July 2011)

Leveraging eDiscovery Technologies for Search and Review

Case: Oracle America v. Google (N.D. Ca. Oct. 20, 2011)

The court ordered Google to produce an email that it previously withheld on attorney client privilege grounds.  While the email’s focus on business negotiations vitiated Google’s claim of privilege, that claim was also undermined by Google’s production of eight earlier drafts of the email.  The drafts were produced because they did not contain addressees or the heading “attorney client privilege,” which the sender later inserted into the final email draft.  Because those details were absent from the earlier drafts, Google’s “electronic scanning mechanisms did not catch those drafts before production.”

The Lesson for 2012: Organizations need to leverage next generation, robust technology to support the document production process in discovery.  Tools such as email analytical software, which can isolate drafts and offer to remove them from production, are needed to address complex production issues.  Other technological capabilities, such as Near Duplicate Identification, can also help identify draft materials and marry them up with finals that have been marked as privileged.  Last but not least, technology assisted review has the potential of enabling one lawyer to efficiently complete the work that previously took thousands of hours.  Finding the budget and doing the research to obtain the right tools for the enterprise should be a priority for organizations in 2012.

Another Must-Read Case: J-M Manufacturing v. McDermott, Will & Emery (CA Super. Jun. 2, 2011)

Conclusion

There were any number of other significant cases from 2011 that could have made this list.  We invite you to share your favorites in the comments section or contact us directly with your feedback.

For more on the cases discussed above, watch this video:

As 2011 comes quickly to a close we’ve attempted, as in years past, to do our best Carnac impersonation and divine the future of eDiscovery.  Some of these predictions may happen more quickly than others, but it’s our sense that all will come to pass in the near future – it’s just a matter of timing.

1. Technology Assisted Review (TAR) Gains Speed.  The area of Technology Assisted Review is very exciting since there are a host of emerging technologies that can help make the review process more efficient, ranging from email threading, concept search, clustering, predictive coding and the like.  There are two fundamental challenges however.  First, the technology doesn’t work in a vacuum, meaning that the workflows need to be properly designed and the users need to make accurate decisions because those judgment calls often are then magnified by the application.  Next, the defensibility of the given approach needs to be well vetted.  While it’s likely not necessary (or practical) to expect a judge to mandate the use of a specific technological approach, it is important for the applied technologies to be reasonable, transparent and auditable since the worst possible outcome would be to have a technology challenged and then find the producing party unable to adequately explain their methodology.
2. The Custodian-Based Collection Model Comes Under Stress. Ever since the days of Zubulake, litigants have focused on “key players” as a proxy for finding relevant information during the eDiscovery process.  Early on, this model worked particularly well in an email-centric environment.  But, as discovery from cloud sources, collaborative worksites (like SharePoint) and other unstructured data repositories continues to become increasingly mainstream, the custodian-oriented collection model will become rapidly outmoded because it will fail to take into account topically-oriented searches.  This trend will be further amplified by the bench’s increasing distrust of manual, custodian-based data collection practices and the presence of better automated search methods, which are particularly valuable for certain types of litigation (e.g., patent disputes, product liability cases).
3. The FRCP Amendment Debate Will Rage On – Unfortunately Without Much Near Term Progress. While it is clear that the eDiscovery preservation duty has become a more complex and risk laden process, it’s not clear that this “pain” is causally related to the FRCP.  In the notes from the Dallas mini-conference, a pending Sedona survey was quoted referencing the fact that preservation challenges were increasing dramatically.  Yet, there isn’t a consensus viewpoint regarding which changes, if any, would help improve the murky problem.  In the near term this means that organizations with significant preservation pains will need to better utilize the rules that are on the books and deploy enabling technologies where possible.
4. Data Hoarding Increasingly Goes Out of Fashion. The war cry of many IT professionals that “storage is cheap” is starting to fall on deaf ears.  Organizations are realizing that the cost of storing information is just the tip of the iceberg when it comes to the litigation risk of having terabytes (and conceivably petabytes) of unstructured, uncategorized and unmanaged electronically stored information (ESI).  This tsunami of information will increasingly become an information liability for organizations that have never deleted a byte of information.  In 2012, more corporations will see the need to clean out their digital houses and will realize that such cleansing (where permitted) is a best practice moving forward.  This applies with equal force to the US government, which has recently mandated such an effort at President Obama’s behest.
5. Information Governance Becomes a Viable Reality.  For several years there’s been an effort to combine the reactive (far right) side of the EDRM with the logically connected proactive (far left) side of the EDRM.  But now, a number of surveys have linked good information governance hygiene with better response times to eDiscovery requests and governmental inquires, as well as a corresponding lower chance of being sanctioned and the ability to turn over less responsive information.  In 2012, enterprises will realize that the litigation use case is just one way to leverage archival and eDiscovery tools, further accelerating adoption.
6. Backup Tapes Will Be Increasingly Seen as a Liability.  Using backup tapes for disaster recovery/business continuity purposes remains a viable business strategy, although backing up to tape will become less prevalent as cloud backup increases.  However, if tapes are kept around longer than necessary (days versus months) then they become a ticking time bomb when a litigation or inquiry event crops up.
7. International eDiscovery/eDisclosure Processes Will Continue to Mature. It’s easy to think of the US as dominating the eDiscovery landscape. While this is gospel for us here in the States, international markets are developing quickly and in many ways are ahead of the US, particularly with regulatory compliance-driven use cases, like the UK Bribery Act 2010.  This fact, coupled with the menagerie of international privacy laws, means we’ll be less Balkanized in our eDiscovery efforts moving forward since we do really need to be thinking and practicing globally.
8. Email Becomes “So 2009” As Social Media Gains Traction. While email has been the eDiscovery darling for the past decade, it’s getting a little long in the tooth.  In the next year, new types of ESI (social media, structured data, loose files, cloud context, mobile device messages, etc.) will cause headaches for a number of enterprises that have been overly email-centric.  Already in 2011, organizations are finding that other sources of ESI like documents/files and structured data are rivaling email in importance for eDiscovery requests, and this trend shows no signs of abating, particularly for regulated industries. This heterogeneous mix of ESI will certainly result in challenges for many companies, with some unlucky ones getting sanctioned because they ignored these emerging data types.
9. Cost Shifting Will Become More Prevalent – Impacting the “American Rule.” For ages, the American Rule held that producing parties had to pay for their production costs, with a few narrow exceptions.  Next year we’ll see even more courts award winning parties their eDiscovery costs under 28 U.S.C. §1920(4) and Rule 54(d)(1) FRCP. Courts are now beginning to consider the services of an eDiscovery vendor as “the 21st Century equivalent of making copies.”
10. Risk Assessment Becomes a Critical Component of eDiscovery. Managing risk is a foundational underpinning for litigators generally, but its role in eDiscovery has been a bit obscure.  Now, with the tremendous statistical insights that are made possible by enabling software technologies, it will become increasingly important for counsel to manage risk by deciding what types of error/precision rates are possible.  This risk analysis is particularly critical for conducting any variety of technology assisted review process since precision, recall and f-measure statistics all require a delicate balance of risk and reward.

(Editor’s note: This special guest post was written by Chitrang Shah, Clearwell Principal Product Manager. He is an RIT alum and avid hiker who works with our engineering team and lead customers to optimize the product for large-scale review. – Kurt)

As we’ve previously shared, our product strategy throughout 2009 and 2010 was to expand the product footprint across the EDRM as customers were demanding a single, end-to-end eDiscovery product. During this period we successfully expanded from our roots in processing, search and analysis to review and production (August 2009), identification and collection (September 2010) and legal hold workflow (March 2011). Over the last several months, our focus has been to go deep in each of these modules and provide features that deliver even greater return on investment to our customers.

Today, I am excited to announce significant new features and feature enhancements to the Clearwell Review and Production Module and say a few words about what motivated us to build these features and how they enable our customers to further streamline their legal review workflow.

There are several exciting features in this release, but I would to like to highlight three in particular:

1. Ability to seamlessly import production load files

Most matters require reviewing relevant documents alongside the documents received from third parties, opposing parties, and even previous litigations. With the new load file import feature, users can now streamline the process of importing load files with three simple steps.

In Step 1, a step-by-step wizard-like interface guides users though the selection of formatting information such as field delimiters and nested value delimiters, metadata information such as bates numbers, family relationships, tags, folders and any number of custom attributes, and content information such as images, extracted text and native files. When the load file has both extracted texts and native files, the wizard gives users an option to specify which content should be used for searching.

In Step 2, the system performs a deep validation of the load file and generates a report documenting any inconsistencies such as missing bates numbers or missing values for required fields found in the load file. As a result, customers have the ability to quickly find and fix any issues with the load file before the import begins.

In Step 3, the system imports the documents and builds analytics. Once this step completes, the imported documents, including all metadata and content, are available for viewing and searching.

All the analytics capabilities customers are familiar with, such as discussion threads and concept search, are also available for documents imported from load files. This allows users to quickly discover documents in the load file that are conceptually similar to natively processed documents, for example.

2. Support for large scale reviews and productions

As the volume of electronically stored information (ESI) continues to grow, our customers find themselves reviewing and exporting more and more documents, and they need a solution that can cope with the massive growth in data. At the same time, they don’t want to spend large sums of money building a server farm in anticipation of the growth. They want the flexibility to add capacity when needed and remove it when not needed.

Clearwell’s scale-out architecture enables administrators to easily add appliances and allocate them to a particular matter and to a specific task using a point-and-click interface.

For example, if an administrator needs to increase the number of reviewers from 200 to 400 in order to meet a tight deadline, he or she can easily add 2 appliances to the cluster and assign them for review. Once the review completes, the administrator can now easily re-assign these appliances for production, allowing users to easily meet deadlines while reducing their overall hardware costs.

This flexibility allows our customers to maximize the use of their hardware resources while providing infinite review, export and production scalability.

3. Streamlined management of exports and productions

Clearwell provides powerful export options, and while our customers use them extensively for creating a variety of different production formats, they typically standardize on a few. Clearwell’s new case export and production templates provide a quick and easy way for case administrators to define the export format once and use it across multiple cases. When exporting documents, users can simply select a template from the list of visible templates in that case. This capability significantly reduces the overhead associated with managing export formats and allows our customers to produce documents in a consistent format across multiple matters.

Additionally, new production pre-mediation reports automatically identify problem documents and group them by issue type for quick resolution. This enables users to preemptively identify and resolve document production issues without delaying entire productions.

Says Wendy Butler Curtis, chair of Orrick, Herrington & Sutcliffe’s eDiscovery Working Group, “Legal review is one of the most challenging phases of the eDiscovery process. As electronic data volumes continue to grow, it is increasingly important to leverage technologies that can streamline and improve legal review, ensure defensibility and reduce costs. Solutions like the Clearwell eDiscovery Platform enable legal teams to create an iterative eDiscovery workflow that allows for more efficient and effective large-scale review.”

We will be showcasing the new features at ILTA (Booth 816) this week in Nashville, so come see us and let us know what you think.

(Chitrang Shah is a Principal Product Manager at Clearwell Systems, now a part of Symantec, and the lead Product Manager for Clearwell’s Processing & Analysis and Review & Production Modules)

Last fall, when Transparent Concept Search was still in development, we showed an early version of it to a group of our customers. Their excitement was palpable, and they spent most of our session together comparing notes about the varied ways they will use it. But at the end of the discussion, one of them asked the question which was on everyone’s mind: “how much will you charge for it?”, or as someone else immediately said “I get charged $200/GB for plain vanilla concept search, so how much of a premium do you think you will get for this?”

Our answer surprised them: there’s no charge. Transparent Concept Search is included in Clearwell for free. Here’s why doing that makes sense:

There are two business strategies in the technology industry which are proven to work. One is to be the low-cost provider and compete on price. These companies, such as Chinese PC manufacturers, do not spend anything on R&D or marketing. Instead, they ruthlessly squeeze out cost savings and pass them on to their customers. The other proven strategy is to be the innovation-leader, whereby you continually delight customers by giving them more and more functionality at the existing price. Players following this strategy are never the cheapest, since they charge a little extra to fund new product development. For example, iPhone is by no means the cheapest smart phone, but its price did not go up when, with the iPhone 4, Apple added video, a forward-facing camera, better battery life, and a retina display.

It is worth noting that either strategy can work, and companies sometimes move between the two, although making that transition is incredibly hard. Staying in the PC industry, Dell started as the low cost provider, but has more recently tried to move up the value chain by investing more in the design of its products. The results, so far, have been mixed.

At Clearwell, our strategy is to be the innovation leader in e-discovery software. We tackle really hard technical problems, solve them in innovative ways, and then seek to delight our users by providing them with breakout, new capabilities at no incremental cost. Transparent Concept Search is a perfect example of this.

Rather than just integrate with concept analysis plug-ins, as pretty much every review platform does, we asked ourselves: if we were to create concept search from scratch specifically for e-discovery, what would we build? As part of that process, we tapped into the latest academic research in semantic analysis coming out of UCLA, University of Pittsburgh, and other universities, and discovered that it offers a solution to the biggest single problem users have with concept search: the heavy computational burden traditional approaches require. By using a variation of the semantic space model which is explained in that new research rather than, say, latent semantic indexing, we can deliver concept searching to much larger legal matters.

Beyond the core technology, we also wanted to change the user experience, by bringing the same level of visibility and control that our users enjoy in keyword search to this domain. Our goal is to enable users to balance both precision and recall in a way that was not previously possible. The result – Transparent Concept Search – is completely seamless within Clearwell in a way that simply cannot be matched by concept search plug-ins to a review platform, which are essentially two separate products from two separate vendors. In summary, it’s a vastly superior user experience – at no incremental cost.

This is the first of many things you will see from us this year. Our team could not be more excited about the new products and ideas that we have in the pipeline.

For years, concept search in electronic discovery has been like concept cars at auto shows: Cool. Slick. The thing that everyone is talking about.

But not ready to move to the assembly line and be put into production.

Like a concept car, concept search has been based on a lot of good ideas and shown a lot of promise. However, it has failed to move beyond a few edge use cases and reach mass adoption in the e-discovery market.  Why is this the case?

It’s not been because it’s an unproven idea or that the basic technology hasn’t been available. In fact, the core algorithm that underlies most existing concept search technologies has actually been around since 1988, when latent semantic analysis (LSA) was first patented by a team from Bell Labs. Over the last 20 years, dozens if not hundreds of companies have sprung up to apply concept search to the broad area of enterprise search and to e-discovery in particular.

To understand why concept search has never taken off, it’s always interesting to look for parallels, and the parallel du jour is social networking. Readers of David Kirkpatrick’s excellent book The Facebook Effect and (perhaps to a lesser, more fictionalized extent) viewers of the movie The Social Network understand that Facebook was far from the first social networking site (remember MySpace? You won’t admit it, but I know you do). But, despite being several years late to the party, Facebook somehow took the core of the social networking idea and presented it to users in a way that really allowed it to “cross the chasm” to the mainstream market.

In introducing Transparent Concept Search, Clearwell plans to help conceptual search cross that same chasm in e-discovery.  In talking to customers over the last couple of years, we have found that there are unmet customer needs with existing concept search products that, once addressed, will really allow its use in e-discovery to flourish – and not just in a way that makes concept search marginally more useful, but, a la Facebook, makes it orders of magnitude more useful.

What are these unmet customer needs?

Ease of use: Historically, concept search has been relatively easy to use in the strictest sense of the word – you type in some terms that represent your concept, and you get a set of search results back, along with some related terms and/or clusters of related documents. Simple, right? The issue is that in most cases that’s not what the user really wants to do. Because concept search is inherently “fuzzy”, users want to be able to refine their concept based on the feedback that they got from their initial search. Concept search, just like keyword search, is an iterative process, and prior-generation technologies have not allowed for that form of iteration. In contrast, Clearwell’s Transparent Concept Search allows concepts to be defined and refined in a way that is intuitive, visual, and (don’t take my word for it, but try it for yourself) fun.

Precision: Traditional concept search increased recall when compared to just keyword search, but it came at the cost of precision. The refinement process facilitated by Clearwell’s Transparent Search addresses this issue by allowing intelligent human input to guide the concept search process. You get the best of both the recall and precision worlds with vastly diminished time and effort.

Defensibility: Even more important than ease of use and precision is defensibility. Defensibility, for those new to the term, isn’t so much about whether the way the algorithms work is known and able to be understood. They are, and aren’t that complicated. Rather, defensibility is about reasonableness: was the concept search a reasonable way of determining which documents are responsive? Without the ability to define your concept in an interactive manner, we believe that the answer has historically been “no”, making concept search nice in theory but unusable in actual legal practice. Transparent Concept Search promises to change that. The end result is a more defensible search process that yields both greater recall and greater precision, enabling users to more quickly analyze case facts, rapidly identify key documents that may have been missed, eliminate irrelevant documents, and prioritize the most relevant documents for review. Clearwell also provides a reporting and auditing feature to document your search, allowing you to improve defensibility by “proving up” what was done.

Low cost: Finally, never underestimate the value of “free” in helping meet the ever-important unmet need of cost predictability and control. Historically, vendors have charged price premiums (often substantial) for concept search. Trying to charge a premium in e-discovery for something that doesn’t fully meet the customer use case and isn’t defensible, and it’s a recipe for low adoption. However, provide a highly useable, effective, and defensible capability as part of the core functionality of today’s leading e-discovery platform, and it starts to look very attractive indeed.

Hopefully you can tell that we’re incredible excited about the promise that this technology holds for the market, and this initial version is really just the beginning. Want to see it for yourself? Check out the video below, visit our web site or, if you are in New York this week, please visit us at LegalTech New York – we would love to see you.

In my last post, I started a discussion on the myths surrounding concept search.  The first myth I dispelled was the “concept search is concept search” myth.  The myth is that there is an agreed upon definition of concept search.  In actuality, when people in electronic discovery use the term concept search, they don’t always mean the same thing.  Frequently they are not actually talking about concept search technology at all and are actually talking about concept or content categorization technology, which is very different.  The second myth that needs dispelling is that concept search is better than keyword search.

The thinking behind this myth goes something like this:

Keyword search has a lot of problems.  It is prone to being over-inclusive, i.e., finding some non-relevant documents, and under-inclusive, i.e., not finding some relevant documents.  Concept search technologies are new and interesting and using these technologies you can find documents that keyword search can’t find.  Therefore, concept search must be better than keyword search.

Let’s examine this thinking.  The first two statements are accurate.  Keyword search is not perfect and can produce over- and under-inclusive results.  And concept search and content categorization technologies can both help identify documents that keyword search technologies might not find.  However, the conclusion that concept search is better than keyword search is not valid and doesn’t follow from these two statements.  Why?

In order to answer this question, we first need to go back to the difference between concept search and content categorization. Because these are different technologies, we really need to separately compare concept search versus keyword search and content categorization versus keyword search.  Let’s start with content categorization and keyword search.

The issue with this comparison is that keyword search and content categorization do different things.  Keyword search can be used in many ways in e-discovery.  The two most common are: (1) analysis or case assessment: finding the hot documents and understanding the matter by determining who knew what, when, how and why, etc., and (2) culling: removing non-responsive documents and/or identifying potentially privileged documents in order to reduce a large, starting set of documents to a smaller set before review.

Content categorization, on the other hand, has historically been used within the review phase of e-discovery.  Categorization can help reviewers to better understand the documents they are reviewing and thus potentially increase the speed of review.  Practitioners with whom I have worked also find that categorization can be useful during analysis by helping to understand a matter and identify potentially important keywords.

However, content categorization has not been used as part of culling.  First, culling needs to be transparent.  You need to be able to get agreement with or at least explain to the opposing side and the court exactly how you have culled the data set.  If you cull based on categories of documents that have been generated by a proprietary, black-box algorithm, it’s going to be difficult to gain agreement on or explain your culling methodology.  This is why the typical method of culling is still to use keyword search and either agree on the set of search terms with the opposing side or to use e-discovery search best practices to perform keyword searches on your own.

Second, content categorization has its own issues when it comes to being over- and under-inclusive.  There is no guarantee that your group of documents that have been categorized as being related to, for example, a company’s hiring policies include all of the documents in your matter related to hiring policies or that they do not include some documents that may not really be related to hiring policies.  Content categorization, like keyword search and virtually every information retrieval technology, is not perfect.

So what about concept search technology?  Surely, concept search technology is better than old, boring keyword search.  Well, actually it’s not that clear-cut.  The problem with concept search technology is that while it might find more relevant documents than plain keyword search, it will also likely find more false positives.  Imagine searching for documents containing “terminate” in an employment matter and your concept search technology automatically searching for “fire”, “dismiss”, etc. as well.  You’ll find more documents related to the termination of employees, but you’ll also find a lot more non-relevant documents concerning house fires, the fire department, etc.

So concept search can help address the under-inclusive problem with keyword search, (though it won’t solve it) and can be helpful during analysis.  But it can often increase the over-inclusive problem.  In addition, today’s concept search technologies share the transparency problem with concept categorization.  These technologies have largely been designed as “black boxes”, which as I have discussed in the past, makes sense for Enterprise search but not for e-discovery search, and, as a result, could also be potentially difficult to explain and defend.   For these reasons, concept search technology isn’t used very much in e-discovery today.  In order for its use to become widespread, it will need to become more transparent.  But that’s a topic for another day.

The bottom line here is that despite all the hype, concept search and content categorization technologies do not solve all the challenges of e-discovery search.  Both of these technologies can be very useful and the technology behind them is always improving.  However, as most of the experienced practitioners I work with already know, these technologies are generally better thought of as supplements to keyword search, not replacements.  The important question is not whether to use one technology over the other but which technology is best suited to your objectives and how best to use all the available technologies to achieve the desired goal.

Concept or content search continues to be a hot topic within the e-discovery community.  There’s a continuous stream of articles that discuss it.  Some that point out the positive.  Others that point out the limitations.  The courts have also gotten involved in the discussion.  Judge Grimm refers to concept search in e-discovery in Victor Stanley, Inc. v. Creative Pipe, Inc., 2008 WL 2221841 (D. Md. May 29, 2008).  Judge Facciola discusses concept search in Disability Rights Council of Greater Washington v. Washington Metropolitan Transit Authority, 242 F.R.D. 139 and other opinions.  Despite (or maybe because of) all the commentary on this topic, I find that while a lot of people think that concept search in e-discovery is good, many are not fully sure of exactly what concept search is, and how it is practically useful in e-discovery.   It’s pretty clear that after several years of commentary and hype, concept search has become something of a buzzword associated with many myths and misconceptions.  In an effort to better understand what concept search is and how it can help in e-discovery, I want to dispel two of the most common myths I have heard.

The “Concept Search is Concept Search” Myth

The first myth around concept search actually revolves around what it is.  In my experience, people tend to lump two different technologies together when talking about concept search: concept search and concept categorization.  It’s very common, for example, to see commentators say concept search even when what they are really talking about is concept categorization.  To make matters more confusing, people also use a plethora of other names including content search, content clustering or concept clustering when what they really mean is concept categorization.

So, what are the differences between concept search and concept categorization?  First, let’s start with concept search.  Concept search technologies find documents containing “concepts”.  I think that the Sedona Conference’s “Best Practices Commentary on the Use of Search & Information Retrieval Methods in E-Discovery“, provides a good definition of “concept” when used in a search context: “the combination of [a] query term and the additional terms identified by the thesaurus.”  In other words, concept search technologies find documents containing a specified term plus additional terms with similar meanings derived from a thesaurus.

Concept categorization, on the other hand, is actually not a search technology at all.  Concept categorization technologies do not “find” documents.  Rather, they categorize or group documents based on their similarity.   There are many different ways to group documents based on similarity.  Techniques include statistical (which assesses similarity based on word frequency), Bayesian classification (which weights words differently depending on factors in addition to statistical frequency, such as where the terms appear in a document), and semantic indexing (which takes into account the fact that many words used in a similar context may have a similar meaning).  It would take more time to describe these technologies in detail but the Sedona commentary has a good summary of these different technologies if you are interested in learning more.

As should now be apparent, these technologies are very different and using the same words to describe them is confusing.  It’s why it’s not surprising that a lot of the users of e-discovery services and software don’t have a strong understanding of what these technologies are or what benefits they can actually provide in practice.  Dispelling the myth that they can be lumped together is a critical first step in any conversation about concept search and how it can help in e-discovery.  This leads us to a second myth, that Concept Search is better than Keyword Search.  I’ll discuss this in my next blog post.