These rulings warn that keyword-based searches prepared without the benefit of "persons qualified to design effective search methodology" frequently do not in a defensible manner isolate potentially relevant and/or privileged information within ESI collections.
There are many good reasons for this, including:
- Understanding the underlying search technology: Different search engines work differently. For example, some will conjugate verbs while most will not. There is variation in how punctuation is handled, and so on.
- Keywords are highly fickle things: They often do not show up where expected, but then turn up in the most inopportune places - in signature lines for example. Keywords alone are often at the same time both under-inclusive (they don't identify everything that is relevant) and over-inclusive (they identify information that is irrelevant).
- Really bad things are hidden under rocks: Rarely will words like "fraud", "embezzlement," or phrases like "tax evasion" or "create a monopoly" actually be found in a hot document. This usually means that a more analytic and empirical approach is needed. For example, many references to various known tax havens such as the Cayman Islands can be an indirect means of identifying potential tax evasion strategies; jargon like "ringfencing business" may be indicative of anti-trust behavior, and so on.
- Multiplicity of channels: People increasingly communicate via various channels, and each channel has different jargon, syntax and vocabulary associated with it. For example, one is likelier to use more abbreviations in an SMS message, and more slang in an instant message than in an email.
- Corporate speak: Organizations have their own internal language such as project code names and acronyms for internal systems and processes.
While the above list is far from exhaustive, it does illustrate some of the problems with relying on keywords without first having your keyword list vetted by experts in legal search. It is especially important to engage expert help before agreeing upon the list with the opposing party.
The most specific guidance was given on May 29, 2008 in Victor Stanley, Inc. v. Creative Pipe, Inc. WL2221841. US Magistrate Judge Paul Grimm provided some questions parties should be prepared to answer regarding their keyword searches:
- What keywords did you use?
- What was the rationale for their selection?
- What were the qualifications of the persons who selected the keywords to be able to design an effective and reliable search and information retrieval method?
- Was the search a simple keyword search or a more sophisticated one that employed, for example, Boolean proximity operators?
- Was there an analysis of the results of the search to assess its reliability, appropriateness for the task, and quality of its implementation?
- Did you follow any industry best standards?
While the focus of Victor Stanley was on privilege waiver, many expect Judge Grimm's test to be widely used regarding defensibility of keyword searches in the future.
The standard approach
Attorneys and case teams historically have created their own search terms using their own analysis of the case, client and witness interviews, assessment of the facts at issue, and their knowledge of typical search terminology (usually, what is allowed in Lexis and Westlaw).
With the growing volume and complexity of ESI, the differences in language from organization to organization, and the advances in search technology, the historical approach has a high likelihood of yielding insufficient or inappropriate search terms, even if prepared by the most skilled of lawyers and case teams. In fact, the most technologically savvy lawyers are beginning to challenge the search methodologies of their opponents in an effort to ensure that they are being given all the relevant facts in discovery.
Our solution
Cataphora employs a team of highly-experienced PhD-level linguists as well as a seasoned team of legal process experts. Our combined expertise (totaling more than 150 years of field experience) is currently offered through the following scenarios:
- Prepare for and participate in meet and confer conferences as they relate to ESI
- Draft discovery requests
- Review and analyze the opposing party's discovery request
- Assist in the preparation of search terms
- Review and analyze the opposing party's suggested search terms
- Provide expert opinions and testimony regarding all of the above
Cataphora has been in the business of providing this level of service for many years.
There is no other company with a comparable track record in using linguistics, statistics and mathematics as they relate to the legal process than Cataphora.
|
|
