Over the last five years I have written about the use of keywords searching in eDiscovery on three separate occasions. There has to be a reason that this one topic has merited this coverage especially as, now, I make it the fourth time.
In October 2015 I wrote an entire post on the subject by way of an introduction to effective keyword searching
Then in April 2016 when considering other eDiscovery features I referred to keyword searching again. Finally, in August 2018, when writing about Technology Assisted Review (TAR) I referred to the increasing doubts on the efficacy of keyword searching alone, within the global industry.
There are many risks. If the list is over inclusive it will defeat the objective of reducing the number of documents for “eyes on” review. If it is under inclusive it is likely to miss potentially important documents. Furthermore, there is nothing conceptual or analytical about keywords, it is merely a way of finding documents that contain the words on the list – remember that Google has been doing this successfully for years!. Specific decided cases in the US have commented upon the dangers of relying on keyword searching – Da Silva Moore v. Publicis Groupe - 287 F.R.D. 182 (S.D.N.Y. 2012) and Nuvasive, Inc. v. Alphatec Holdings, Inc. No. 18-CV-0347 (S.D. Ca.) (10/7/2019) to name but two.
I was motivated to once again re-visit the topic following a recent article from Australia .
This article asks, rhetorically, if 2020 will see the end of keyword searching in favour of more advanced analytics such as technology assisted review (TAR). I will come right off the fence here and say that in South Africa, a fledgling eDiscovery jurisdiction, the use of keywords still has a valuable place. People here are still learning the value of eDiscovery technology and because it is not mandatory within our Rules there are fewer cases in which it is being employed. There are a number of investigations, rather than litigation, in which eDiscovery technology is being used and for sure keywords are helping investigators find specific documents based upon what they are trying to achieve.
I have mentioned before that what is crucial is to work with experts on your list of keywords who can advise as to the efficacy of the list and also not to assume that the first list is the only list – it needs to be re-visited, probably more than once. Of course, if the volumes are large enough the best practice is to combine keyword searching with other features of eDiscovery technology involving analytics. My concern is that lawyers and service providers alike are insufficiently experienced as yet in SA to do this, which highlights my previous comments about choosing your service provider wisely. Bluntly, there are very few individuals in SA who have the experience and knowledge to give firm and proper advice on this. Therefore, the temptation is to pay “token” homage to eDiscovery technology by simply throwing a list of keywords at a database and saying you are an expert.
A little while ago I published a post with my thoughts on eDiscovery and the State Capture Commission or Zondo Commission as it is often referred to. How delighted I am to read the latest news concerning the President’s decision to amend the rules of the Commission so that evidence gathered by and for it can be used for other agencies such as NPA, Hawks etc..
The gazetted amendment will also allow the NPA to access staff who are familiar and have been working on the various investigations within the Commission. I am sure that these include investigators and eDiscovery experts all of whom have been bound by agreements of secrecy, thus far, for obvious reasons. It is what I referred to in my post in that the country, if not the world, needs to see positive results arising from the evidence brought to the Commission. It is particularly pleasing to see this even before DCJ Zondo has considered, let alone published, his findings. As my interest is the increased use of eDiscovery technology (quite apart from my interest like everyone else in SA to see justice arising out of corruption allegations) I am very happy, as it is clear that the NPA etc. will continue to use the technology that has already played its part in the Enquiry. There will be some challenges ahead in terms of migrating the appropriate data especially when it will be important to include results of some of the filtering work already done. Once again, eDiscovery technology experts will be required to save time and money and add valuable knowledge. All good news from where I am sitting.
I conclude this post with an update on the eDiscovery book for South Africa that I am co-writing with Ismail Hussain SC as so many of you have asked me about it and commented upon it. I can tell you that we are very very close to sending the manuscript to Lexis Nexis for editing and publication. We are all motivated to having this book available as soon as ever possible and definitely this year.