Mining Commonsense Facts from the Physical World
Abstract
Textual descriptions of the physical world implicitly mention commonsense facts, while the commonsense knowledge bases explicitly represent such facts as triples. Compared to dramatically increased text data, the coverage of existing knowledge bases is far away from completion. Most of the prior studies on populating knowledge bases mainly focus on Freebase. To automatically complete commonsense knowledge bases to improve their coverage is under-explored. In this paper, we propose a new task of mining commonsense facts from the raw text that describes the physical world. We build an effective new model that fuses information from both sequence text and existing knowledge base resource. Then we create two large annotated datasets each with approximate 200k instances for commonsense knowledge base completion. Empirical results demonstrate that our model significantly outperforms baselines.
Keywords
Cite
@article{arxiv.2002.03149,
title = {Mining Commonsense Facts from the Physical World},
author = {Yanyan Zou and Wei Lu and Xu Sun},
journal= {arXiv preprint arXiv:2002.03149},
year = {2020}
}
Comments
The experiment part is not insufficient which might confuses the readers, while we are not planning to improve it for now. To ensure the quality of arxiv papers, we would like to withdraw this submission