Myth and Reality about Predictive Coding
May 23, 2012
The myth of a technology can often differ from its reality, especially in one that is still evolving. Predictive coding is currently experiencing controversy regarding its cost versus efficiency according to Will Predictive Coding Live Up to the eDiscovery Hype?.
Predictive Coding has been hailed a technology which provides lower costs with lighter burdens. The unfortunate results that deem it myth fall to the process necessary for it to function. It is an evolving technology.
Its reality is evolving as:
“With the promise of transparency and simpler workflows, predictive coding technology should eventually live up to its billing of helping organizations discover their information in an efficient, cost effective and defensible manner. As for now, the “promise” of first generation predictive coding tools appears to be nothing more than that, leaving organizations looking like the cash-strapped “Monopoly man,” wondering where there litigation dollars have gone.”
Reality is that Predictive Coding can’t exist without human’s providing the data, and then the program optimizes it. The process combines people, technology and workflow to find documents referencing keywords. The three basic components are:
- To predict utilizing predictive analytics.
- To code, utilizing a keyword to locate relevant documents.
- To process a proven workflow.
Improvements are possible given the technological advances in the industry. This new technique has potential and may yet evolve into a functional, efficient means to acquire data. For now, Predictive Coding is stuck in between myth and reality.
Jennifer Shockley, May 23, 2012
Comments
2 Responses to “Myth and Reality about Predictive Coding”
Jennifer, I agree with the weaknesses of predictive coding.
The key to predictive coding is still keywords – the input. Bad input generates bad output. Therefore, the predictive coding successes I have heard of involve situations where a considerable amount of review has already been completed, or a particular “smoking gun” is already in possession.
Key word search is a tremendous technology and has room for improvement (i.e. generation 2 of predictive coding), but the only way to create a paradigm shift in this industry is to look outside of search, which few are willing to do.
Catelas begins with behaviors and relationships and is able to generate a tremendous amount of case intelligence in 30 minutes without the use of a single key word. Add in key words, and a full case report is generated in 3-4 hours – a report that would take a week (or a month) to generate utilizing other tools. The report represents a “cliff’s notes” – the complete abridged story about the case.
Understanding people relationships enable the keyword technologies to focus on fewer communications and generate fewer false positives.
Catelas’ Early Case Intelligence Report provides tremendous inputs for predictive coding, but … if counsel already has “smoking guns” and a clear understanding of what the outcome will be — then she will move towards case resolution, and not spend more money on technology.
Ben Hogan
Catelas, Inc.
The resulting documents can
[…] Bowes that specializes in eDiscovery is Content Analyst. Content Analyst is one of the leaders in predictive coding technology and auto categorization. Litigation support companies adding services such as ECA and […]