Truth-tracking with Non-expert Information Sources
Authors: Joseph Singleton, Richard Booth
JAIR 2024 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We study what can be learned when receiving propositional reports from multiple nonexpert information sources. We investigate the extent to which truth-tracking is possible, and describe what information can be learned even if the actual world cannot be pinned down uniquely. This paper adapts learning-theoretic notions from formal learning theory and in particular its intersection with belief revision (Baltag et al., 2019) to handle non-expert information sources. We establish the limits of learning in this setting, and conditions under which one can learn the true facts of the world as well as the true extent of the expertise of the sources. We go on to characterise truth-tracking learning methods in terms of syntactic postulates, and look specifically at some methods previously introduced by Singleton and Booth (2022). |
| Researcher Affiliation | Academia | Joseph Singleton EMAIL Richard Booth EMAIL Cardiff University, UK |
| Pseudocode | No | The paper describes methods and characterizations using logical frameworks and postulates, but does not include any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any statements about open-source code availability, nor does it include links to repositories or mention code in supplementary materials. |
| Open Datasets | No | The paper describes a theoretical framework and uses conceptual examples (e.g., a medical scenario, finite sets of propositional variables and cases) rather than empirical datasets. There is no mention of publicly available or open datasets used for experiments. |
| Dataset Splits | No | The paper is theoretical and does not conduct experiments on datasets, therefore, there is no information regarding training/test/validation dataset splits. |
| Hardware Specification | No | The paper focuses on theoretical aspects and does not describe any experimental setup that would require specific hardware specifications. |
| Software Dependencies | No | The paper is theoretical and does not describe any specific software implementations or dependencies with version numbers. |
| Experiment Setup | No | The paper is theoretical and does not describe any experimental setup, hyperparameters, or system-level training settings. |