Keyword Searching For E-disclosure And Digital Forensics

By Author: Luciano Castaneda
Total Articles: 150
Comment this article

In the vast majority of e-disclosure and also digital forensic exercises case study is driven usually by keyword searching. There are numerous discussions within the field of e-disclosure about Technologies Assisted Review (Black) (that is predictive coding) taking out the industry’s reliance on keyword looking. Over the next couple of years Tar residue will be used on some projects, but key phrases will still be the main driver of search and review.

As keywords and phrases are such a essential part of an e-disclosure exercise, it is vital to get these right. Selecting a key phrase set is often more of an art than a research. It is rarely easy to predict how many visits a keyword will yield, however an awareness of how keywords function and the profile with the data you are seeking will help to avoid key phrases that may yield a lot of documents to review likely.
Below are some quick tips to follow when choosing search term sets:

Take time to comprehend the data you are looking - Avoid generally occurring words or phrases. What may be a good keyword for one dataset may not be for an additional.

Be ...
... as certain as possible without being restricted - The more certain the keywords, the greater. However, overly-restrictive keywords could lead to key documents being missed, so an equilibrium is required.

Obtain indicative hit counts before beginning review - equipment will allow reporting on the number of hits to get a keyword before launching the documents with regard to review. Use this performance and adjust keywords and phrases if the hit quantities are unexpectedly high or low.

Avoid large key phrase sets - The greater keywords you run, the greater documents for evaluation. Large keyword sets may be appealing to create an all-encompassing population regarding documents for assessment, however you may find that population is unwieldy.

Make use of wildcards - A wildcard is really a character that can be substituted with other characters. For instance a search for ‘Brib*’ would go back ‘Bribery’, ‘Bribe’, ‘Bribes’, etc. These can help to accommodate for spelling and tense variants.

Use Boolean operators to target documents responsive to multiple keyword - Files containing ‘payment’ may not be relevant but documents containing ‘secret’ AND ‘payments’ may be. Additional operators include: NOT, OR, WITHIN as well as XOR.
Group keywords into sets - Either by priority or even by issue, it will help focus the assessment process.

Be aware of ‘standard text’ - Check what terms and terms exist in commonly occurring e mail footers and boilerplate terms and conditions. Choose keywords that do not exist in these or all of the documents that contain the standard text will be presented regarding review. Some evaluation platforms, such as Clearwell, permit you to exclude blocks regarding text such as email footers from the search, enabling you to use keywords which exist in footers.

http://www.cclgroupltd.com/e-disclosure/
http://www.cclgroupltd.com/e-disclosure/collections/