Spreadsheet Datasets

We have introduced two spreadsheet datasets:

• SAUS: The 2010 Statistical Abstract of the United States (SAUS) consists of 1,369 spreadsheet files totaling 70MB. We downloaded the dataset from the U.S. Census Bureau. It covers a variety of topics of general public interest, such as state-level finances, educational attainment, levels of public health, and so on. [Download]


• WEB: Our Web dataset consists of 410,554 Microsoft Excel files from 51,252 distinct Internet domains. They total 101 GB. We found the spreadsheets by looking for Excel-style file endings among the roughly 10 billion URLs in the ClueWeb09 Web crawl. [Download]