Issue link: https://epubs.iltanet.org/i/13683
Even after a well-designed data minimization (“landscaping”) strategy is employed, there are typically still a significant number of documents to be reviewed. For many review teams, dealing with large numbers of documents requires brute force tactics. This means increasing the number of reviewers on a project without making changes to workflow or to strategies that were probably designed for smaller teams and document collections. However, text analytics search technology enables smarter review strategies that improve both the speed and quality of the expensive review process. This search technology, which evaluates unstructured text, enables users to automatically group documents into conceptually related subsets. Document searches can also be based on concepts. By using text analytics to structure a review, litigation professionals can significantly reduce discovery costs by increasing document review speeds. BASIC LINEAR REVIEW VERSUS USING TEXT ANALYTICS Case teams navigate data sets in a linear review by reading documents in order. This order is often by date or by document control number. Reviewers move through the collection of documents one at a time, identifying relevant documents that are interspersed with non-responsive material. This type of standard linear organization presents challenges that can impact both the speed and tenacity of the review. For example, a reviewer may be responsible for reading thousands of e-mail messages and documents from one custodian, knowing that the majority of the documents are non-responsive. Since the material is organized by date, the subject matter of each document in sequence can vary greatly. This forces the reviewer to examine each document for potential relevance, thus increasing review time and costs. Text analytics allows reviewers to navigate data sets by organizing documents into “conceptual order.” This helps separate sets of relevant documents from sets of potentially non-responsive documents, and eliminates the speed and quality issues caused by the mixing of relevant and non-responsive documents throughout the collection. Reviewers are able to gain review momentum by applying knowledge gained from each document to the entire document set. This is akin to a mountain biker muscling through the beginning climb and then reaping the benefits on the downhill journey. In addition, a reviewer has the ability to gain insight into a custodian’s involvement well before completing the review of an entire document set. The review of one conceptually similar document group could be all that is needed to form a strong opinion of the facts, or at least to allow for more informed decision making. USING CONCEPT SEARCH AND CLASSIFICATION TO ACCELERATE REVIEW Text analytics features, as applied to document review in litigation, fall into two broad categories: concept search and classification. Both concept search and classification can be applied in different ways to improve and accelerate review. CONCEPT SEARCH To execute a concept search, a user enters text (e.g., a few sentences or a paragraph from a www.iltanet.org Case/Matter Management 25