Equivio recently introduced Equivio>Relevance™, an expert-guided system that enhances the eDiscovery process through automated document prioritization. Traditionally, people, such as attorneys, engaged in intense investigations use keywords to pre-filter documents prior to detailed review. According to the TREC (Text Retrieval Conference) studies, the keywords method tends to miss most of the relevant documents, while yielding mainly irrelevant documents. Equivio>Relevance is designed to address these limitations, introducing a higher level of flexibility, control and accuracy into the eDiscovery process.

I have been interested in the various forms that search can take for some time so I was eager to learn about this new product. Recently, I spoke with Warwick Sharp at Equivio to get the details. He said that TREC studies have shown that keyword search only finds 20 -30 % of relevant documents and returns 70 -80 % junk that needs to be weeded out.

Equivio>Relevance takes a different approach. You train it on what documents are relevant using successive groups, each of about 40 documents. The system learns the characteristics of relevant documents. The system monitors its performance relative to the human expert. When it reaches the desired training threshold it shifts from learning mode to analysis mode and rates the complete set of documents.

The keywords approach is typically binary, meaning that documents are either included or not. There is no graduated scale of relevance. This does not allow for relative ranking of documents, making it difficult to manage and prioritize document review. With Equivio>Relevance documents are rated on a 0 – 100 relevance score. You can then choose the level of relevance to include in your review. For example, in the sample we looked at, the top 9% of documents sorted by Equivio>Relevance contained 70% of the relevant documents, and the top 20% of documents contained 90% of the relevant documents.  There is a nice interactive graph that visualizes the sample to let you more easily make decisions about what level you want to explore.

Equivio>Relevance also helps you find what you did not know you were looking for. It generates a list of keywords that characterize relevant documents in the collection. These automatically generated keywords can be used to supplement and enhance the manual list of keywords built by the legal team.

The efficiency flexibility offered by Equivio>Relevance is especially useful in today’s tighter economy. Law firms can no longer work on maximizing billable hours. They are being handed budgets for investigations. Equivio>Relevance lets you decide on the level of effort to match the budget.

Equivio>Relevance can be implemented in either standalone mode or, using an SDK, can be tightly integrated with the customer’s processing and database environment.  The product supports the processing of both native and extracted files. It was first released on a private basis this January and its public release occurred in June.

I think this is a very creative alternative to keyword search and there should be many applications beyond legal firms.