Human-in-the-Loop Disinformation Detection: Stance, Sentiment, or Something Else?

TL;DR

This study compares sentiment analysis, aspect-based sentiment analysis, and stance detection to determine the most effective technique for human-in-the-loop disinformation detection across diverse topics and datasets.

Contribution

It evaluates and compares the effectiveness of three NLP techniques for disinformation detection in a human-in-the-loop setting using multiple COVID-19 datasets.

Findings

Stance detection outperforms sentiment analysis in disinformation detection.

All techniques show reduced accuracy on unseen topics.

Qualitative insights suggest stance detection is most suitable for practical applications.

Abstract

Both politics and pandemics have recently provided ample motivation for the development of machine learning-enabled disinformation (a.k.a. fake news) detection algorithms. Existing literature has focused primarily on the fully-automated case, but the resulting techniques cannot reliably detect disinformation on the varied topics, sources, and time scales required for military applications. By leveraging an already-available analyst as a human-in-the-loop, however, the canonical machine learning techniques of sentiment analysis, aspect-based sentiment analysis, and stance detection become plausible methods to use for a partially-automated disinformation detection system. This paper aims to determine which of these techniques is best suited for this purpose and how each technique might best be used towards this end. Training datasets of the same size and nearly identical neural architectures (a BERT transformer as a word embedder with a single feed-forward layer thereafter) are used for each approach, which are then tested on sentiment- and stance-specific datasets to establish a baseline of how well each method can be used to do the other tasks. Four different datasets relating to COVID-19 disinformation are used to test the ability of each technique to detect disinformation on a topic that did not appear in the training data set. Quantitative and qualitative results from these tests are then used to provide insight into how best to employ these techniques in practice.