Extracting Qualitative Causal Structure with Transformer-Based NLP

TL;DR

This paper introduces a transformer-based NLP method that extracts variables, causal relationships, and qualifiers from text, enabling understanding of complex qualitative causal structures across various sources.

Contribution

It presents a novel transformer architecture that jointly identifies variables, causal links, and qualifiers from natural language, advancing qualitative causal analysis in NLP.

Findings

Effective extraction of causal relationships from diverse texts

Promising results in academic, news, and social media data

Improved understanding of qualitative causal structures

Abstract

Qualitative causal relationships compactly express the direction, dependency, temporal constraints, and monotonicity constraints of discrete or continuous interactions in the world. In everyday or academic language, we may express interactions between quantities (e.g., sleep decreases stress), between discrete events or entities (e.g., a protein inhibits another protein's transcription), or between intentional or functional factors (e.g., hospital patients pray to relieve their pain). This paper presents a transformer-based NLP architecture that jointly identifies and extracts (1) variables or factors described in language, (2) qualitative causal relationships over these variables, and (3) qualifiers and magnitudes that constrain these causal relationships. We demonstrate this approach and include promising results from in two use cases, processing textual inputs from academic publications, news articles, and social media.