enron_xls
Spreadsheets from the Enron Corpus
Enron Spreadsheet Corpus
The Enron Corpus is a massive database of emails amassed in the investigation of the former Enron Corporation. The original corpus is available as a series of PST email archives. The emails include tens of thousands of spreadsheets.
Various sources point to the existence of a version of the dataset with all attachments. Those links have been removed but the dataset was preserved on the Internet Archive
The original dataset included personally identifiable information such as birth dates and Social Security numbers. Nuix produced a cleaner dataset and made it available to the community. Unfortunately this dataset was removed.
The spreadsheets in this dataset are in their original format, including BIFF2, TSV, semicolon-delimited values, SYLK, and HTML files saved as XLS. They have been de-duplicated by MD5 hash.
Files
Methodology
EDRM Dataset
The included
The actual dataset is a series of ZIP files. It is possible to cherry-pick from the ZIP archives without having to download the entire 74 GB archive.
Nuix Dataset
Starting from the cleaned email set, each PST file was downloaded and processed
using the excellent pst-extractor Node module.
Every available XLS attachment (these emails predate XLSX, which was introduced in 2007) was extracted, and the files were de-duplicated based on MD5 checksum.
Any duplicates of files in the EDRM set were removed.