Get the Most Out of eDiscovery Analytics:
Hosting analytics and technology-assisted review have transformed the eDiscovery industry. However, there are other options to help perform faster and more efficient reviews. Keyword expansion, concept searching, and clustering are just a few of those options.
By using algorithms, it is easy to identify words in close proximity to key term(s). Using keyword expansion, can uncover additional relevant emails or documents that could have otherwise been missed.
Concept Searching and “Find Similar”
If a search term has multiple meanings, submitting a sentence or paragraph as a concept search can be more accurate than using keywords alone. After a concept search is run, batching out the results prior to batching out the rest of the data set can help reveal key documents faster. Using “Find Similar” on privileged search terms also aids in identifying additional terms that should be run through the privilege filter. Concept searching also helps with accurately spot checking the work of reviews and locates any privilege level documents that were missed after the first-level review.
It is important to weed out non-relevant material in large data sets. Identifying the most interesting clusters to focus on is also an essential strategy. Clustering enables reviewed to focus on similar documents in a consecutive manner for a faster review. Locating large groups of email or documents that contains non-responsive material enables them to be tagged quickly without wasting valuable review time.
Email Threading and Near Duplicate Detection
Reviewers can focus on only inclusive emails and make decisions about other emails in that chain by using email threading. This can reduce the overall review volume. Email providers can render emails differently which prevents them from being removed by traditional deduplication methods. Near duplicate detection addresses this issue by locating different versions of document and grouping the versions together.
For more Tidbits & Thoughts, please click here.