Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Explainable Fact-checking through Question Answering

Jing Yang, Didier Vega-Oliveros, Taís Seibt, Anderson Rocha

TL;DR

This work addresses fact-checking explainability through question answering, and proposes an answer comparison model with an attention mechanism attached to each question that can achieve state-of-the-art performance while providing reasonable explainable capabilities.

Abstract

Misleading or false information has been creating chaos in some places around the world. To mitigate this issue, many researchers have proposed automated fact-checking methods to fight the spread of fake news. However, most methods cannot explain the reasoning behind their decisions, failing to build trust between machines and humans using such technology. Trust is essential for fact-checking to be applied in the real world. Here, we address fact-checking explainability through question answering. In particular, we propose generating questions and answers from claims and answering the same questions from evidence. We also propose an answer comparison model with an attention mechanism attached to each question. Leveraging question answering as a proxy, we break down automated fact-checking into several steps -- this separation aids models' explainability as it allows for more detailed analysis of their decision-making processes. Experimental results show that the proposed model can achieve state-of-the-art performance while providing reasonable explainable capabilities.

Explainable Fact-checking through Question Answering

TL;DR

This work addresses fact-checking explainability through question answering, and proposes an answer comparison model with an attention mechanism attached to each question that can achieve state-of-the-art performance while providing reasonable explainable capabilities.

Abstract

Misleading or false information has been creating chaos in some places around the world. To mitigate this issue, many researchers have proposed automated fact-checking methods to fight the spread of fake news. However, most methods cannot explain the reasoning behind their decisions, failing to build trust between machines and humans using such technology. Trust is essential for fact-checking to be applied in the real world. Here, we address fact-checking explainability through question answering. In particular, we propose generating questions and answers from claims and answering the same questions from evidence. We also propose an answer comparison model with an attention mechanism attached to each question. Leveraging question answering as a proxy, we break down automated fact-checking into several steps -- this separation aids models' explainability as it allows for more detailed analysis of their decision-making processes. Experimental results show that the proposed model can achieve state-of-the-art performance while providing reasonable explainable capabilities.

Paper Structure

This paper contains 14 sections, 1 equation, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Proposed Methodology
  3. Question and Answer Generation
  4. Answer Pair Comparison
  5. Experimental Setup
  6. Dataset
  7. Implementation details
  8. Baselines
  9. Results and Analysis
  10. Comparison with baselines
  11. Attention visualization
  12. Ablation study
  13. Limitations
  14. Conclusion

Figures (3)

  • Figure 1: Answer comparison model with attention on questions. $C$ represents a given claim, $Q_i$ represents $i_{th}$ questions, and $({A^C}_i, {A^E}_i)$ represents $i_{th}$ answer pairs for claim and evidence. $n$ denotes the number of questions and answer pairs.
  • Figure 2: An example of our model generated questions, answer pairs, and attention weights. The question with the highest weight is in bold, and the second highest is underlined.
  • Figure 3: Answer comparison model without attention on questions. $C$ represents a given claim, $Q_i$ represents $i_{th}$ questions, and $({A^C}_i, {A^E}_i)$ represents $i_{th}$ answer pairs for claim and evidence. $n$ denotes the number of questions for claim $C$.