Relevance Ranking is the Key to Early Case Assessment

This article considers the question of what is the best technology for doing Early Case Assessment (ECA), which is briefly defined by the author as the process of analyzing a subset of case-related information in order to determine whether to pursue the case. The author also addresses the question of whether it makes more sense to purchase or subscribe to ECA software.

The central challenge in ECA, the author says, is to identify the highest value materials as expeditiously as possible so the most informed decisions can be made. Keyword matching, formerly the default strategy, produces a flood of irrelevant documents and misses relevant ones, while failing to prioritize what it does produce. Other techniques include “sentiment detection,” which looks for apparently emotion-laden communications, and social network analysis, which give weight to who the communicators are and how they relate to each other. Both techniques, while useful, have serious shortcomings.

Relevance ranking is an important recent advance. A knowledgeable attorney repeatedly ranks small sets of documents, each time interacting with the software, so the software effectively “learns” to apply the attorney’s judgment to the larger set. What may be a two-day development process, the authors says, results in a method that will handle millions of documents daily.

In cases where the decision is to litigate, ECA can facilitate setting an appropriate e-discovery budget, and properly done it can either avert or justify millions of dollars in legal costs.