We explore methods to extract relations between named entities from free text in an unsupervised setting. In addition to standard feature extraction, we develop a novel method to re-weight word embeddings. We alleviate the problem of features sparsity using an individual feature reduction. Our approach exhibits a significant improvement by 5.8% over the state-of-the-art relation clustering scoring a F1-score of 0.416 on the NYT-FB dataset.
@article{arxiv.1801.07174,
title = {Unsupervised Open Relation Extraction},
author = {Hady Elsahar and Elena Demidova and Simon Gottschalk and Christophe Gravier and Frederique Laforest},
journal= {arXiv preprint arXiv:1801.07174},
year = {2018}
}