cheb's Introduction

CHEB-dataset

Release 1, 2018-03-23

This is data used in "Concept Hierarchy Extraction from Textbooks" (Wang et al., 2015)

Data Description

This dataset contains two parts:

original book content and wikipedia dictionary as described in the paper
key concepts label

For book content and wiki dictionary:

A wiki dictionary {bookname}.wikis is collected by running a BFS crawler on wiki anchor network (as described in the paper)
A clean wiki dictioanry {bookname}.wikis_clean is generated by filtering categories that has too few pages in the dictionary (as describe in section3.1 in the paper)
Each file under folder {bookname}_content contains content for each book chapter (each chapter in a seperate file)

For key concepts label:

{bookname}_vote_final.csv (separated by semicolons)
col0: chapter number
col1: chapter title
col2: if a concept is a key concept in this chapter; If this column is "1" or "2", this wiki concept is a key concept; Otherwise, it is not

Citation

Please cite the following paper if you use this data.

@inproceedings{wang2015concept,
  title={Concept hierarchy extraction from textbooks},
  author={Wang, Shuting and Liang, Chen and Wu, Zhaohui and Williams, Kyle and Pursel, Bart and Brautigam, Benjamin and Saul, Sherwyn and Williams, Hannah and Bowen, Kyle and Giles, C Lee},
  booktitle={Proceedings of the 2015 ACM Symposium on Document Engineering},
  pages={147--156},
  year={2015},
  organization={ACM}
}

If you have any problems, please contact Shuting Wang at [email protected].