About
ReVerb is a program that automatically identifies and extracts binary relationships from English sentences. ReVerb is designed for Web-scale information extraction, where the target relations cannot be specified in advance and speed is important.
To get a better idea of what ReVerb does:
- Download the code on github.
- Read Identifying Relations for Open Information Extraction.
- Browse ReVerb's extractions from 500 million Web pages.
Code
ReVerb is released under an academic license. For instructions on how to run ReVerb or use it in your own code, please see the README file (also included in the download).
- ReVerb 1.3 (on github)
- Previous Versions: 1.0
- README
- LICENSE
Data
A collection of 15 million ReVerb extractions is available for academic use. The extractions are the result of running ReVerb on the ClueWeb09 dataset. Please review the information in the README_data.txt. Also, consider downloading the smaller sample of Wikipedia-only extractions before downloading the full set.
- README_data.txt
- CHANGES_data.txt
- 100 Sample Extractions
- ReVerb Wikipedia Extractions 1.1 (28MB, compressed)
- ReVerb ClueWeb09 Extractions 1.1 (826MB, compressed)
Labeled data from the experiments in the EMNLP 2011 paper is available here.
Credits
ReVerb was developed by the following people at the University of Washington's Turing Center as part of the KnowItAll Project:
- Anthony Fader
- Michael Schmitz
- Robert Bart
- Stephen Soderland
- Oren Etzioni
ReVerb uses the following code and data, which are included in the release:
- OpenNLP, including trained models
- Weka
- Guava
- Apache Commons
- A subset of pages from Wikipedia
Contact
Please send all questions, comments, and bugs to the the ReVerb mailing list.