Multi-Hop Fact Checking of Political Claims

TL;DR

This paper introduces PolitiHop, a new dataset for complex multi-hop fact checking of political claims, and explores transfer learning techniques to improve reasoning over interconnected evidence.

Contribution

It creates a novel annotated dataset for multi-hop political claim verification and evaluates transfer learning methods to enhance reasoning performance.

Findings

Multi-hop reasoning is complex and challenging.

In-domain transfer learning improves fact-checking accuracy.

Specialized architecture modeling evidence reasoning yields best results.

Abstract

Recent work has proposed multi-hop models and datasets for studying complex natural language reasoning. One notable task requiring multi-hop reasoning is fact checking, where a set of connected evidence pieces leads to the final verdict of a claim. However, existing datasets either do not provide annotations for gold evidence pages, or the only dataset which does (FEVER) mostly consists of claims which can be fact-checked with simple reasoning and is constructed artificially. Here, we study more complex claim verification of naturally occurring claims with multiple hops over interconnected evidence chunks. We: 1) construct a small annotated dataset, PolitiHop, of evidence sentences for claim verification; 2) compare it to existing multi-hop datasets; and 3) study how to transfer knowledge from more extensive in- and out-of-domain resources to PolitiHop. We find that the task is complex and achieve the best performance with an architecture that specifically models reasoning over evidence pieces in combination with in-domain transfer learning.