Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Can Factual Statements be Deceptive? The DeFaBel Corpus of Belief-based Deception

Aswathy Velutharambath, Amelie Wührl, Roman Klinger

TL;DR

This paper addresses how deception relates to factuality and personal belief in argumentation by introducing DeFaBel, a German corpus of belief-based deception. It crowdsources argumentative texts and annotates them for deception, belief, and factuality, labeling deception when authors argue in opposition to their own beliefs, enabling disentanglement of these intertwined factors. The corpus construction combines belief distribution assessment, distribution-based filtering, and postgeneration annotations, yielding 1031 German texts (643 deceptive, 388 non-deceptive) from 164 participants. Key findings show higher confidence and topic familiarity when arguments align with belief, while arguing for factual statements lowers self-reported persuasiveness; the work lays the groundwork for deception-aware fact checking and more robust belief-aware NLP tools. This resource and analysis advance understanding of how belief and factuality shape deceptive and persuasive language, with practical implications for detecting deception and improving fact checking in multilingual settings, particularly German.

Abstract

If a person firmly believes in a non-factual statement, such as "The Earth is flat", and argues in its favor, there is no inherent intention to deceive. As the argumentation stems from genuine belief, it may be unlikely to exhibit the linguistic properties associated with deception or lying. This interplay of factuality, personal belief, and intent to deceive remains an understudied area. Disentangling the influence of these variables in argumentation is crucial to gain a better understanding of the linguistic properties attributed to each of them. To study the relation between deception and factuality, based on belief, we present the DeFaBel corpus, a crowd-sourced resource of belief-based deception. To create this corpus, we devise a study in which participants are instructed to write arguments supporting statements like "eating watermelon seeds can cause indigestion", regardless of its factual accuracy or their personal beliefs about the statement. In addition to the generation task, we ask them to disclose their belief about the statement. The collected instances are labelled as deceptive if the arguments are in contradiction to the participants' personal beliefs. Each instance in the corpus is thus annotated (or implicitly labelled) with personal beliefs of the author, factuality of the statement, and the intended deceptiveness. The DeFaBel corpus contains 1031 texts in German, out of which 643 are deceptive and 388 are non-deceptive. It is the first publicly available corpus for studying deception in German. In our analysis, we find that people are more confident in the persuasiveness of their arguments when the statement is aligned with their belief, but surprisingly less confident when they are generating arguments in favor of facts. The DeFaBel corpus can be obtained from https://www.ims.uni-stuttgart.de/data/defabel

Can Factual Statements be Deceptive? The DeFaBel Corpus of Belief-based Deception

TL;DR

This paper addresses how deception relates to factuality and personal belief in argumentation by introducing DeFaBel, a German corpus of belief-based deception. It crowdsources argumentative texts and annotates them for deception, belief, and factuality, labeling deception when authors argue in opposition to their own beliefs, enabling disentanglement of these intertwined factors. The corpus construction combines belief distribution assessment, distribution-based filtering, and postgeneration annotations, yielding 1031 German texts (643 deceptive, 388 non-deceptive) from 164 participants. Key findings show higher confidence and topic familiarity when arguments align with belief, while arguing for factual statements lowers self-reported persuasiveness; the work lays the groundwork for deception-aware fact checking and more robust belief-aware NLP tools. This resource and analysis advance understanding of how belief and factuality shape deceptive and persuasive language, with practical implications for detecting deception and improving fact checking in multilingual settings, particularly German.

Abstract

If a person firmly believes in a non-factual statement, such as "The Earth is flat", and argues in its favor, there is no inherent intention to deceive. As the argumentation stems from genuine belief, it may be unlikely to exhibit the linguistic properties associated with deception or lying. This interplay of factuality, personal belief, and intent to deceive remains an understudied area. Disentangling the influence of these variables in argumentation is crucial to gain a better understanding of the linguistic properties attributed to each of them. To study the relation between deception and factuality, based on belief, we present the DeFaBel corpus, a crowd-sourced resource of belief-based deception. To create this corpus, we devise a study in which participants are instructed to write arguments supporting statements like "eating watermelon seeds can cause indigestion", regardless of its factual accuracy or their personal beliefs about the statement. In addition to the generation task, we ask them to disclose their belief about the statement. The collected instances are labelled as deceptive if the arguments are in contradiction to the participants' personal beliefs. Each instance in the corpus is thus annotated (or implicitly labelled) with personal beliefs of the author, factuality of the statement, and the intended deceptiveness. The DeFaBel corpus contains 1031 texts in German, out of which 643 are deceptive and 388 are non-deceptive. It is the first publicly available corpus for studying deception in German. In our analysis, we find that people are more confident in the persuasiveness of their arguments when the statement is aligned with their belief, but surprisingly less confident when they are generating arguments in favor of facts. The DeFaBel corpus can be obtained from https://www.ims.uni-stuttgart.de/data/defabel
Paper Structure (25 sections, 1 equation, 6 figures, 6 tables)

This paper contains 25 sections, 1 equation, 6 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Deception
  4. Factuality
  5. Belief in Persuasive Argumentation
  6. Corpus Creation
  7. Question Selection
  8. Belief Distribution Assessment
  9. Argument Generation
  10. Distribution-based Filtering.
  11. Instruction Formulation.
  12. Argumentative Text Generation.
  13. Crowd-sourcing Study Details
  14. Data Analysis
  15. Overview.
  16. ...and 10 more sections

Figures (6)

  • Figure 1: Deception label assignment based on author's belief and factuality of the statement.
  • Figure 2: The corpus creation process.
  • Figure 3: Distributions of confidence regarding persuasiveness, familiarity with the topic and belief in the prompting statement.
  • Figure 4: Belief assessment question as displayed in Google Forms
  • Figure 5: Example of attention check question
  • ...and 1 more figures