Incorporating Commonsense Knowledge Graph in Pretrained Models for Social Commonsense Tasks

TL;DR

This paper introduces methods to incorporate external commonsense knowledge graphs into pretrained language models to improve their performance on social commonsense reasoning tasks, demonstrating effectiveness in various data regimes.

Contribution

It proposes two novel approaches for integrating commonsense knowledge graphs into pretrained models, enhancing social intelligence in NLP tasks.

Findings

Improved performance on SocialIQA with the proposed methods.

Effective in both limited and full training data scenarios.

Demonstrates the benefit of external knowledge integration for social reasoning.

Abstract

Pretrained language models have excelled at many NLP tasks recently; however, their social intelligence is still unsatisfactory. To enable this, machines need to have a more general understanding of our complicated world and develop the ability to perform commonsense reasoning besides fitting the specific downstream tasks. External commonsense knowledge graphs (KGs), such as ConceptNet, provide rich information about words and their relationships. Thus, towards general commonsense learning, we propose two approaches to \emph{implicitly} and \emph{explicitly} infuse such KGs into pretrained language models. We demonstrate our proposed methods perform well on SocialIQA, a social commonsense reasoning task, in both limited and full training data regimes.