Counterfactual Thinking for Long tailed Information Extraction

Information Extraction (IE) aims to extract structured information from unstructured texts . In practice, the long-tailed and imbalanced data may lead to severe bias issues for deep learning models . Existing works are mainly from computer vision society, leveraging re-balancing, decoupling, transfer learning and causal inference to address this problem . However, these approaches may not achieve good performance on textual data, which involves complex language structures . To this end, we propose a novel framework (named Counterfactual-IE) based on language structure and causal reasoning with three key ingredients . We propose a flexible debiasing approach for more robust prediction during the inference stage . Experimental results on three IE tasks across five public datasets show that our model significantly outperforms the state-of-the-art models by a large margin in terms of by a huge margin in the model . We also discuss some interesting findings based on our observations. We also talk about our model. We hope to use this model to improve our findings on the ACE2005 dataset to improve the accuracy of the IE tasks. It also discusses some interesting results based on their findings. They also propose a model that is based on an explicit language structure to better calculate the

Links: PDF - Abstract

Code :

None

Keywords : model - based - language - propose - information -

Leave a Reply

Your email address will not be published. Required fields are marked *