Some methods adopt multi-instance learning by combining sentences with same entity pairs and then selecting informative instances from them.

Selecting Informative Instances

Incorporating extra context information to denoise DS data has also been explored, such as incorporating KGs as external information to guide instance selection and adopting multi-lingual corpora for the information consistency and complementarity.

Incorporating Extra Context Information

 Many methods tend to utilize sophisticated mechanisms and training strategies to enhance distantly supervised NRE models.  For example, combine different architectures and training strategies to construct hybrid frameworks, incorporate a soft-label scheme by changing unconfident labels during training.

Sophisticated Mechanisms and Training Strategies 

The existing methods to alleviate the noise problem can be divided into three major approaches: selecting informative instances, incorporating extra context information, and utilize sophisticated training strategies.

Hamilton College

Supervised NRE models suffer from the lack of large-scale high-quality training data, since manually labeling data is time-consuming and human-intensive. To alleviate this issue, distant supervision (DS) assumption has been used to automatically label data by aligning existing KGs with plain text.

Utilizing More Data: Methods

Han, X. (n.d.). More Data, More Relations, More Context and More Openness:
A Review and Outlook for Relation Extraction *, Retrieved June 4, 2022, from https://arxiv.org/pdf/2004.03186.pdf

Learn Before

Related

Learn After