[author: Mallory Maier-Acheson]

While many litigation departments are familiar with structured analytics such as e-mail threading and near duplicate analysis, a powerful supplement that need not be forgotten is conceptual analytics.  Conceptual analytics workflows can efficiently help to illuminate what the data contains, assist with quality control, narrow in on key documents, prioritize review, and even reduce review of non-relevant documents.  Below are four processes that fall under the umbrella of conceptual analytics and are key to daily litigation workflows.

Opposing Party Production Analysis – Efficient Review is Key

Opposing party productions can be time consuming and expensive to review. We find that implementing intelligent processes to reduce the number of documents reviewed and prioritize documents of concern ultimately increases productivity and is key to the analysis.

When faced with multiple productions, grouping conceptually-similar documents along with utilizing the outbound coding fields to categorize the in-bound production has proven efficient and effective.  Leveraging the prior coding fields allows for prioritization based on issue and key categorization, generating those pressing documents to review first. 

Another method often leveraged is to review the in-bound production using an active learning or continuous active learning tool, whereby the analytics engine actively queues up for review the documents deemed to be the highest relevantly ranked set based on current coding decisions. 

Categorization – Use Example Documents to Group Similar Documents Together for Coding

Along with using categorization for opposing party production analysis, this tool can also be leveraged to identify similar key documents for deposition preparation, quality control (“QC”) privilege review on out-bound productions, and prioritize review based on responsive, key, or issue coding.

Clustering – Immediately Identify Concepts and Similarly Grouped Documents

With no previous coding necessary, clustering will bring forward the conceptual topics contained within a data set while grouping conceptually similar documents together.  This tool is especially helpful for gaining knowledge in unfamiliar data sets, identifying problematic data sets, and prioritizing certain clusters of documents by promoting documents in the same clusters.  Combining this tool with any previous coding, such as key or privileged documents, can aid in prioritizing review of any documents contained within similar clusters.

Keyword Expansion – Language for All

Not everyone uses the same vernacular.  A great tool to aid with this language difference is keyword expansion, as it helps attorneys to investigate the language of the custodians.  This process is used with known keywords to expand search terms or keyword lists and identify code-words and atypical uses.

Prioritization, reduced review, quality control, targeted identification, and investigation all under one umbrella – now that deserves an award!