CMVC+: A Multi-View Clustering Framework for Open Knowledge Base Canonicalization via Contrastive Learning

Source code and data sets for the CMVC+: A Multi-View Clustering Framework for Open Knowledge Base Canonicalization via Contrastive Learning

Dependencies

• Compatible with Python 3.6
• Dependencies can be installed using requirements.txt
• Please download the init_dict folder from this webpage: https://drive.google.com/file/d/1PNMSO5ud5UuArU9vrtwTvBzECn4Pfudr/view?usp=sharing
• Please download the crawl-300d-2M.vec.zip from https://fasttext.cc/docs/en/english-vectors.html into init_dict.

Data sets

• Please download the ReVerb45K and OPIEC59K data sets from this webpage:  https://drive.google.com/file/d/1vetosoVBf89-It1cD671AfiP0ZrcBz-P/view?usp=sharing

• Please download the NYTimes2018 data set from the webpage of its original authors:  https://heathersherry.github.io/ICDE2019_data.html

Usage

Run the main code:

• python Main_CMVC_plus_ReVerb45K.py
• python Main_CMVC_plus_NYTimes2018.py
• python Main_CMVC_plus_OPIEC59K.py

Contact

Yang Yang (yang.yang@insight-centre.org)

(back to top)