Migrated from eDJGroupInc.com. Author: Greg Buckles. Published: 2010-07-26 10:44:09Format, images and links may no longer function correctly. As some of you might recall, I briefly ran Attenex’s services group before the potential FTI acquisition reminded me how much I liked to work for myself. I always liked Document Mapper, aka ‘the Petri Dish’, and wanted to expand the usage of their conceptual clustering, social networking and other visual analytics beyond the review interface. The FTI Technology team has accomplished this by combining three products under the Ringtail Analytics brand. Even better, they are breaking free of the volume based licensing that has dominated the market for far too long. Corporations can purchase these add-on Ringtail modules via a traditional enterprise license. The idea is to apply a variety of analytics on the early assessment samples and against collections as large as a million items to support ECA, search creation, processing parameters and to optimize automated review sets management.

FTI Technology - Document Mapper

Document Mapper aka 'The Petri Dish'

So how is this supposed to actually work? It starts with their Smart Decision tool that extracts metadata and other characteristics into an OLAP cube data structure (also called a Hypercube or Datacube). This is a fancy way of building a relational database with metrics and other facet dimensions built in to make reporting and navigation much more efficient. Think of it as the backbone to support the kinds of guided search that I explored in my last article.  The dynamic tabular interface reminds me of the new Relativity Pivot feature, but adding in full text facets like keyword statistics by custodian, source, type, etc. Whether you are using this for the early key custodian assessment collection or on the large preservation collection, the ability to generate these kinds of interactive reports linked to live data without paying per GB are what corporate litsupport have been looking for.

FTI Technology - Smart Decision OLAP cube

Smart Decision OLAP cube

Following the Smart Decision tool is a new concept visualization interface called Concept Mapper that extracts concepts and keywords from your raw data collection to enable the power user to organize, filter, prioritize and cluster prior to actual review in Document Mapper. The individual document cluster visualization of Document Mapper was really optimized for sets of 1,000-5,000 items. You could get useful information out of sets up to 20,000 items, but it was not practical to try to view entire collections. The new visualization is built for collections up to 1 million items and allows the power user to control the conceptual confidence thresholds so that you can easily identify clusters of high confidence documents to route into review queues. The tool includes some sophisticated automated review routing, quality control and management functionality. These features seem to be aimed at the kinds of very large reviews still managed by firms and specialized outsourced providers.

FTI Technology - Concept Mapper

Concept Mapper

Although I am disappointed that the current Ringtail Analytics module does not include the new Acuity suggested coding functionality, it does add the traditional Document Mapper to Ringtail’s linear review platform. This is a good fit for the typical small scale litigation with 3-5 primary custodians and a single paralegal or inside counsel doing the review. The Acuity predictive coding propagates decisions from a high confidence training set down to multiple reviewers in the classic large review project. The FTI Technology team has taken aim at the corporate identification, processing and review market with this offering. I hope to give you feedback from a real world implementation of these tools in the future to see how they handle large and complex data sets. It is good to see renewed innovation and offerings from early pioneers like Attenex and Stratify. It is an indicator that acquisitions do not have to kill sharp products.

0 0 votes
Article Rating