The workweek is the best time to start a family -- A Study of GPT-2 Based Claim Generation

TL;DR

This paper presents a GPT-2 based pipeline for generating coherent claims, analyzes their types and truthfulness, and examines how claim generation and retrieval can enhance each other in social media contexts.

Contribution

It introduces a novel GPT-2 based approach for claim generation, evaluates claim quality and veracity, and explores the synergy between claim generation and retrieval tasks.

Findings

GPT-2 can generate coherent claims with varying types and truthfulness.

Claim retrieval and generation tasks can mutually improve each other's performance.

The approach offers insights into social media information dissemination.

Abstract

Argument generation is a challenging task whose research is timely considering its potential impact on social media and the dissemination of information. Here we suggest a pipeline based on GPT-2 for generating coherent claims, and explore the types of claims that it produces, and their veracity, using an array of manual and automatic assessments. In addition, we explore the interplay between this task and the task of Claim Retrieval, showing how they can complement one another.