Hey folks – I\u2019m working on a research project around eDiscovery workflows and ran into a gap with the datasets that are publicly available.<\/p>\n
Most of the \u201copen\u201d collections (like the EDRM Micro Dataset) are useful for testing parsers because they include many file types – Word, PDF, Excel, emails, images, even forensic images – but they don\u2019t reflect how discovery actually feels<\/em>. They\u2019re kinda just random files thrown together, without a coherent story or links across documents.<\/p>\n What I\u2019m looking for is closer to a realistic \u201cmock case\u201d dataset: I\u2019ve checked Enron, EDRM, and RECAP, but those either don’t have narrative structure or aren’t really raw discovery.<\/p>\n Does anyone know of (preferably free and public): Thanks in advance!<\/p>\n<\/div>\n submitted by \/u\/darkprime140 <\/a>
\u2022 A set of documents (emails, contracts, memos, reports, exhibits) that tell a narrative when read together (even if hidden in a large volume of files)
\u2022 Something that could be used to test workflows like chronology building, fact-mapping, or privilege review
\u2022 Public, demo, or teaching datasets are fine (real or synthetic)<\/p>\n
\u2022 Law school teaching sets for eDiscovery classes
\u2022 Vendor demo\/training corpora (Relativity, Everlaw, Exterro, etc.)
\u2022 Any academic or professional groups sharing narrative-style discovery corpora<\/p>\n
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
