CO-NNECT: A Framework for Revealing Commonsense Knowledge Paths as Explicitations of Implicit Knowledge in Texts

TL;DR

This paper introduces Co-nnect, a framework that uses combined language models and knowledge graphs to generate explicit knowledge paths, revealing implicit connections between sentences in texts.

Contribution

It presents a novel joint framework leveraging finetuned language models and knowledge graphs to generate and evaluate explicit knowledge paths for implicit sentence connections.

Findings

Generated paths are meaningful and high-quality.

The framework effectively reveals implicit knowledge in argumentative texts.

Evaluation shows improved path quality over baseline methods.

Abstract

In this work we leverage commonsense knowledge in form of knowledge paths to establish connections between sentences, as a form of explicitation of implicit knowledge. Such connections can be direct (singlehop paths) or require intermediate concepts (multihop paths). To construct such paths we combine two model types in a joint framework we call Co-nnect: a relation classifier that predicts direct connections between concepts; and a target prediction model that generates target or intermediate concepts given a source concept and a relation, which we use to construct multihop paths. Unlike prior work that relies exclusively on static knowledge sources, we leverage language models finetuned on knowledge stored in ConceptNet, to dynamically generate knowledge paths, as explanations of implicit knowledge that connects sentences in texts. As a central contribution we design manual and automatic evaluation settings for assessing the quality of the generated paths. We conduct evaluations on two argumentative datasets and show that a combination of the two model types generates meaningful, high-quality knowledge paths between sentences that reveal implicit knowledge conveyed in text.