Applied Artificial Intelligence in Litigation and Compliance: Is it Really Ready for Prime Time?
Over the years, information management disciplines have evolved through several discrete stages. What began centuries ago as rudimentary manual document indexing systems, have developed into complex taxonomical hierarchies and classifications that afford individuals and organizations the ability to classify massive amounts of data and information with extraordinary degrees of granularity. The most commonly used of these systems have followed a natural path across various industries and disciplines, in large part because they make sense. They take into account the most basic elements of the natural creation, aggregation, storage, use, retrieval and disposition of information.
Much of the methodology which forms the foundation of what is colloquially referred to as Artificial Intelligence (AI), originated years ago. Arguably, the fathers of AI could be considered to be the mathematicians Sir Thomas Bayes and George Boole. These gentlemen are responsible for Bayesean algorithms and Boolean logic. They were prolific contributors to statistics, probability and logic theories. Today there are viewed retrospectively by many as the founders of computer science.
Skipping the nauseating description of the informational evolutionary tree, we segue to June, 2006 to discuss Artificial Intelligence in the context of litigation and compliance. In 2001: A Space Odyssey, Hal, the omnipresent sentient computer, was constantly parsing, running calculations and analyzing what at the time, would have been considered unthinkable amounts of data. The result of Hal's computations were outputted as information in a form from which immediate qualitative and substantive decisions can be made and actions taken. This is effectively the state of the technology today.
It is not uncommon for complex litigation document productions to generate multiple terabytes of responsive data from which relevant evidence must be culled. To whittle data collections down to manageable size, there are a variety of methodologies. For years target searching comprised of Boolean, fuzzy, stem, proximity searching has formed the basis of common textual analysis. However concept searching, much more closely related to how we as humans make an association between evidence and issues, has evolved and matured to a point where its practical application in the context of litigation can generate a significant ROI. The next most important question is, "How? Show me an example of how it helps me in managing this litigation or conducting the risk assessment in this merger, acquisition or divestiture for the purposes of due diligence."
While I could pose several answers to this question myself, I've chosen instead to invite the industry leading litigators and technologists, and you, the avid lover of the practice and litigation support disciplines, to weigh in and comment.
"Remember...that which you know, irrespective of your role as the prosecution or the defense, is a lot more valuable than that which you don't, so try to learn as much as you can about what you don't know. - Perry Mason
Much of the methodology which forms the foundation of what is colloquially referred to as Artificial Intelligence (AI), originated years ago. Arguably, the fathers of AI could be considered to be the mathematicians Sir Thomas Bayes and George Boole. These gentlemen are responsible for Bayesean algorithms and Boolean logic. They were prolific contributors to statistics, probability and logic theories. Today there are viewed retrospectively by many as the founders of computer science.
Skipping the nauseating description of the informational evolutionary tree, we segue to June, 2006 to discuss Artificial Intelligence in the context of litigation and compliance. In 2001: A Space Odyssey, Hal, the omnipresent sentient computer, was constantly parsing, running calculations and analyzing what at the time, would have been considered unthinkable amounts of data. The result of Hal's computations were outputted as information in a form from which immediate qualitative and substantive decisions can be made and actions taken. This is effectively the state of the technology today.
It is not uncommon for complex litigation document productions to generate multiple terabytes of responsive data from which relevant evidence must be culled. To whittle data collections down to manageable size, there are a variety of methodologies. For years target searching comprised of Boolean, fuzzy, stem, proximity searching has formed the basis of common textual analysis. However concept searching, much more closely related to how we as humans make an association between evidence and issues, has evolved and matured to a point where its practical application in the context of litigation can generate a significant ROI. The next most important question is, "How? Show me an example of how it helps me in managing this litigation or conducting the risk assessment in this merger, acquisition or divestiture for the purposes of due diligence."
While I could pose several answers to this question myself, I've chosen instead to invite the industry leading litigators and technologists, and you, the avid lover of the practice and litigation support disciplines, to weigh in and comment.
"Remember...that which you know, irrespective of your role as the prosecution or the defense, is a lot more valuable than that which you don't, so try to learn as much as you can about what you don't know. - Perry Mason
posted by Richard E. Davis, JD at 6:37 AM
0 comments
links to this post
