The document aligned data set contains two files:

normal.txt
simple.txt

with English Wikipedia sentences and Simple English Wikipedia sentences respectively.  The
files are article aligned and represent the same ~60K articles topics.  The data set contains
all Simple English Wikipedia articles that have a corresponding article in English Wikipedia.

The Wikipedia data was downloaded in May 2011.

The data set contains metadata about where each sentence came from.  Each line in the files are 
of the form:

article_title<TAB>paragraph_number<TAB>sentence

Additional information about the data can be found at:

http://www.cs.middlebury.edu/~dkauchak/simplification/

For questions/concerns, please contact David Kauchak at Middlebury College.

To cite, please use:
Improving Text Simplification Language Modeling Using Unsimplified Text Data. David Kauchak (2013). In Proceedings of ACL.
