ZyLAB offers visual classification for eDiscovery

ZyLAB, a provider of eDiscovery and Information Management solutions, has announced that it has incorporated true native visual search and categorisation in its eDiscovery and Information Management solution to enable the speedy and accurate identification of non-textual information. The inclusion of Visual Classification together with ZyLAB’s Audio Search, completes ZyLAB’s multi-media search capabilities.

Enterprise Big Data contains large volumes of electronically stored information (ESI) that is non-textual, for example: pictures, video and audio. Processing, searching and classification of these types of ESI without textual information in the image or in the image related documents adds significant costs and risks to all parties involved in an eDiscovery process. 

ZyLAB’s Visual Classification automatically recognises pictures and identifies amongst others: people, babies, elderly people, flowers, cars, planes, in- and outdoor scenes, and many other concepts. The new functionality is perfectly usable for the identification of images of personal identifiable information (PII), potential intellectual property, handwritten notes, check’s, ID’s, and other information that otherwise cannot be recognised automatically.

Professor Johannes Scholtes, Chief Strategy Officer of ZyLAB: “Under the Federal Rules of Civil Procedure (FRCP) all information, including all forms of non-text based data, is subject to eDiscovery. But if such an image does not contain textual information, automatic Optical Character Recognition (OCR) or any other text-based search and technology assisted review tool is useless. Until now, an expensive and long manual review process was the only alternative.” 

ZyLAB is the first vendor to market this technology in the eDiscovery space. Using Visual Classification during the legal processing of data, images are handled by the image classification engine. This makes it possible to tag images and video with the available concepts. Based on this tagging, users can determine the workflow for processing. One can decide not to OCR images that contain landscapes, babies or flowers (for example) saving significantly on processing time as performing OCR on images can be time-consuming, especially when you perform a four-way OCR.

The native visual search and categorisation technology that ZyLAB now offers as a new add-on to the ZyLAB eDiscovery and Production Platform is traditionally used in law enforcement and intelligence applications such as video surveillance and picture classification. 

Prof. Scholtes: “The cost and time savings for reviewers is paramount. This technology enables reviewers to exclude all non-relevant images, instantly find all images that contain handwritten notes for further investigation, and quickly and effectively locate potentially confidential information such as copies of credit cards and ID’s.

Together with ZyLAB’s Audio Search, ZyLAB’s new Visual Classification now completes ZyLAB’s multi-media search capabilities. In combination with ZyLAB’s award winning global text-search, text-mining  technology, and support for more than 450 languages, 1000 electronic file formats, and 200 different repository types, including the cloud, ZyLAB will be able to find more relevant information in the eDiscovery market, regardless of its form, shape, language or location.